Yesterday’s Moon: A Decipherment of the Classic Mayan Adverb ak’biiy

by David Stuart (University of Texas at Austin)

Deciphering the Adverb Ak’biiy, “Yesterday”

Maya inscriptions contain a few terms or phrases that we can classify as temporal adverbs, helping to specify the timing of events relative to the text’s internal time-frame. One such term is sahm-iiy, spelled sa-mi-ya, “earlier today,” which I identified some years ago in two moon age records at Palenque (Figure 1) (first presented in Houston, Robertson and Stuart 2000). In this context, before the verb hul-iiy, sahm-iiy simply states that the new moon appeared only within a day of some notable event in the narrative “present.” The full phrase illustrated here can be translated as sahm-iiy hul-iiy, “earlier today it arrived.” Its suffix -iiy is the Classic Mayan form traceable to proto-Mayan *-eer, “ago, before,” and is an extremely common deictic suffix found on most if not all of these adverbs that mark a point in the past. It can appear on intransitive verbs, adverbs, as well as on some enumerated nouns. Grammatically, sahm-iiy and its relatives work in a way similar to standard day-counts that reckon a span of time from some earlier event to up to a present one. For example, we find in other lunar day-counts the expressions jo’lahuun-ih-iiy (15-ji-ya),  “fifteen days ago…”, or wukbix-iiy (7-bi-xi-ya), “seven days ago.”

sahmiiy glyph
Figure 1. Parallel examples of sahm-iiy hul-iiy, “earlier today it arrived.” (a) PAL: Palace Tablet, Q10-R10, (b) PAL:T.XXI bench edge. Drawings by D. Stuart.

Here I identify another temporal pronoun that I read as ak’b-iiy, “yesterday,” or “the night before.” Like the examples just cited, these occur in records of moon ages, where short-term temporal expressions of less than thirty days are routinely found. The first instance comes from Stela F at Qurigua (Figure 2a), as part of the moon age record accompanying the Long Count 9.16.10.0.0 1 Ahau 3 Zip (March 14, 761 CE). Here before the verb hul-iiy (hu-li-ya), “it arrived,” (Macleod 1990), we find a glyph consisting of a turkey’s head with the suffix sign ya. Infixed into the turkey is a bi syllable. The second case comes from Zoomorph O’ (Figure 2b), where the same glyph appears but now with the prefix a-, also before hul-iiy. These related glyphs have remained undeciphered until now, but they have generally been recognized as indicating new moon, or the start of the lunar month.

ak'biiy glyphs
Figure 2. Two examples of ak’b-iiy hul-iiy, “yesterday it arrived.” (a) QRG Stela F, F6, (b) QRG: Zoo O’, J1-I2. Drawings by D. Stuart

There is good evidence to show that the turkey head is read AK’, based on the noun ak’ or ak’ach, “turkey hen.” In the inscriptions of La Corona, the very same sign appears in the spelling of the personal name Chak Ak’ Paat Kuy, where it alternates with the syllabic comination a-k’a (Figure 3) (Houston, Stuart and Zender 2017, Stuart and Zender 2018). The same turkey sign also occurs on Stela 1 of Dos Pilas in a variant of the “dance” verb AK’-ta-ja (Figure 4b) (see Grube 1990), suggesting it may be a head variant or a graphic elaboration of the more abstracted AK’ sign we find more frequently in that position (we will return to the connection between these signs a little further on).

Ckak ak'
Figure 3. The AK’ turkey logogram in a name from La Corona. (a) CRN: Elem. , (b) CRN:Elem. 56, pB1, (c) CRN: Elem. 56, pF2. Drawings by D. Stuart
ak'taj verbs
Figure 4. Substitution of AK’ in AK’-ta-ja, ak’t-aj, “he dances” (a) DPL: Stela 11, (b) DPL: Stela 1. Drawings by D. Stuart.

On Quirigua Stela F we therefore have a plausible reading AK’-bi-ya for the glyph before hu-li-ya. On Zoomorph O’ we have the very same expression, but with an a added as a prefix on AK’. The bi infix again looks to be present (photos are murky), so the full form here seem to be a-AK’-bi-ya. There can be little doubt that these two glyphs spell the temporal adverb ak’b-iiy, a form found in Ch’olan languages meaning “yesterday,” or “last night,” based upon the noun ahk’ab, “night.” Note its use in these sentences from from Ch’orti and Ch’ol:

ak’bi patneen, yesterday I worked (Wisdom 1950)
ac’bi tsa’ huliyon ilayi, yesterday I arrived here (Aulie and Aulie 1978)

In the last example cited, it is interesting to see that Ch’ol ac’bi serves as an adverb before a derived form of hul, “to arrive,” much as we find in the case of the Lunar Series examples from Quirigua. The same term can be traced more widely throughout lowland Mayan languages (all shown in their original orthographies).

Ch’olti’: acbihi, yesterday (Moran 1935)
Ch’orti’: ak’bi, yesterday, of yesterday (Wisdom 1950)
Ch’ol: ’ak’-b’i, yesterday; ayer (Hopkins, et al 2008)
Ch’ol: ac’bi, ayer (Aulie and Aulie 1978)
Chontal: ?äk’-bi, yesterday, before (Knowles 1984)
Chontal: äc’-bi, ayer (Keller and Luciano 1997)
Tzeltal: ahkab-ey, anoche (Polian 2020)
Tzotzil: ak’ub-e, anoche (Kaufman and Justeson 2003)
Yukatek: ak’be’, anoche, la noche anterior (Barrera Vásquez 1980)

In the two contexts from Quirigua the full verbal expression is therefore ak’biiy huliiy, “yesterday it arrived,” the subject being the lunar month of 29 or 30 days. This might be roughly understood to saying that the moon is simply one day old. However, we should exert some caution in assuming so, and reflect further on the descriptive language the Maya used in such records. The moon’s “arrival” is not simply the astronomical new moon, which corresponds to its dark, invisible phase. “Arrival” should be understood as referring to the moon’s first visibility, as a thin waxing crescent in the darkened sky. Landa made this point in his Relación, wherein he states that “they counted (the lunar month) from the time at which the new moon appeared until  it no longer appeared”  (Tozzer 1941:133). This is confirmed, I believe, by the form of the HUL logogram used in the vast majority of moon age records, which depicts a hand pointing at the crescent (see Houston 2012). The “pointing at the moon” form of HUL is especially clear in its Early Classic examples (Figure 5). All of this is to say that first visibility would fall a two or three days (depending on timing) after astronomical new moon. Thus ak’biiy huliiy can be best analyzed as an explicit statement about first visibility happening “yesterday,” falling two or even three days after the astronomical new moon.

HUL signs
Figure 5. Selection of Early Classic HUL logograms, emphasizing its origin as a hand point at the moon. (a) RAZ: Tomb 1, (b) COP: St. 63, (c) CLK: Celt, (d) PNG: St. 30, (e) RSB: HS 3. Drawings by D. Stuart.
ak'biiy Coba
Figure 6. Alternate spelling of ak’biiy huliiy from COB: Panel D. Drawing by D. Stuart.

Apart from Quirigua, there may be one other example of ak’biiy in a text from Coba. This is Panel D, the unusual rectangular panel or altar with a spiral-shaped inscription. This text opens with a Lunar Series, probably a continuation of a calendrical text now lost. The initial glyph of the text (Figure 6) is “Glyph D” of the Lunar Series, incorporating the verb hul-iiy (HUL-li-ya). Before this we have a glyph shows an infixed bi element and a ya suffix. I suggest the main sign here is the alternate logogram for AK’, known from the frequent dance verbs mention above (AK’-ta-ja) (see Figure 4a). At first I wondered if this could be a spelling of bix-iiy, using a logogram for BIX that I identified in 1996 (Stuart 2012). However, it is possible to distinguish that form from the somewhat similar AK’, which consistently shows two darkened elements along the internal curved line. BIX seems to regularly feature a single darkened element, with an infixed bi. I suspect that this AK’ is graphically related to the turkey head variant, perhaps originating as a pars pro toto of it.

The use of the turkey AK’ in these spellings brings up a couple of interesting points regarding hieroglyphic orthography. First, we have here the rare use of a logogram for purely phonetic purposes (the adverb ak’biiy having nothing to do with turkeys). It also reflects a high degree of phonetic sensitivity in Maya script. As we have seen, the derived form ak’b-iiy results from two morphophonemic process that work on the underlying noun ahk’ab, “night.” The first is vowel syncope. In Ch’olan languages, any stem of more than two syllables, such as that formed by the combination of ahk’ab and -iiy, sees the loss of its penultimate vowel, in this case a (Kaufman and Norman 1984:86). The second process is the loss of the h before a cluster of two consonants. The root for “turkey” is ak’, lacking the internal h we find in ahk’ab. As Marc Zender has pointed out to me (personal communication, 2020), a spelling such as a-AK’-bi-ya appears to be a remarkably precise means of representing the phonetic result of these standard processes. A similar situation presumably exists in the spellings of the verb for “dance,” usually spelled AK’-ta-ja (see Grube 1990), at times with the very same AK’ turkey sign. The proto-Ch’olan verb root is ahk’ot, and the addition of the intransitivizing suffix -aj necessitates the same two processes just described, the result being ak’t-aj, “(s)he dances.”

Thus far I have not encountered the temporal adverb ak’biiy outside of the context of moon age records. This is not terribly surprising, given how Lunar Series passages focus on short-term time frames involving less than thirty days, sometimes focused on time changes within a single day, as in sahm-iiy.

Moon Ages and Correlations

The ak’biiy huliiy statements may be significant in considering the astronomical correlations of certain Maya calendar dates. For example, the Long Count on Quirigua, Stela F is 9.16.10.0.0 1 Ahau 3 Zip, falling on March 11, 761 according to the 584283 correlation. This is astronomical new moon, which occurred in the pre-dawn hours of that day.

YEAR 0761
New Moon          First Quarter       Full Moon         Last Quarter
Mar 11 05:43      Mar 18 01:19        Mar 25 04:59   Apr 2 05:49
(dates and times in Universal Time minus 6 hours).

It should be emphasized that astronomical new moon describes a phase of complete invisibility before the first sliver of the crescent is visible. Yet we know in Maya terms that the moon’s “arrival” was its first appearance to the naked eye, as discussed above, and as indicated by the visual forms of several HUL logograms (Prager 2020).  It is very difficult to see the young waxing crescent of the moon within 24 hours of astronomical new moon, in fact, suggesting that visibility with the naked eye during the night spanning March 11 and 12 would have been extremely unlikely. The night of March 13 is a more likely time for the moon’s true “arrival” and visibility. If we align these lunar phenomena with the different correlation constants for 9.16.10.0.0, we have:

584283, night of March 11, 761 – Astronomical new moon (invisible)
584284, night of March 12, 761 – initial waxing crescent, minimal visibility
584285, night of March 13, 761 – waxing crescent, newly visible
584286, night of March 14, 761 – one day after visibility

Here we see that the moon record on Stela F, explicitly stating that the moon “arrived yesterday,” accords best with either the 584285 (March 13) or perhaps even more so with the 584286 correlation (March 14) proposed by Martin and Skidmore (2012).

The ak’biiy huliiy date on Zoomorph O’ is the accession of the ruler we know as “Sky Xul,” on 9.17.14.16.18 9 Edznab 1 Kankin. In the 584283 correlation this falls on October 9, 785. New moon, the period of invisibility, had occurred just before midnight on October 7, into October 8, and would have continued for another 24 hours. First visibility would most likely have been no earlier than the night of October 10 or 11.

YEAR 0785
New Moon          First Quarter       Full Moon         Last Quarter
Oct 7 23:02        Oct 15 17:35         Oct 22 08:42    Oct 29 12:08
(dates and times in Universal Time minus 6 hours).

Using the 584285 constant, the accession date is October 11, and with the 584286 it falls on October 12. The ak’biiy statement on Zoomorph O’ therefore is in keeping with either of these correlations, but not so much with the 584283. Of the two, the 584286 may even seem more fitting, marking the arrival of the moon, or first visibility, on October 11. Zoomorph G celebrated the Period ending 9.17.15.0.0 5 Ahau 3 Muan, only 22 days after the accession of Sky Xul. Its moon age is recorded is recorded as 23 days, precisely what we would expect if we reckon from Zoomorph O’ and its ak’biiy statement of the moon being first visible on October 11. It is worth noting that these two uses of ak’biiy at Quirigua also correlate well with the probable eclipse record at Santa Elena Poco Uinic Stela (July 16, 790), which seems best anchored in the 584286 correlation, as discussed by Martin and Skidmore (2012).

These are only cursory observations about the correlation issue, and far more thought needs to go into these questions. The main point to stress is that, until recently, Maya epigraphers simply classified the two glyphs we can now read as ak’biiy and sahmiiy as general indicators of “new moon.” Now we can be more precise about their meanings. One is “earlier today,” and another is “yesterday.” The implications of these decipherments should be pondered further, for they might help in fine-tuning the correlation of the ancient Maya calendar with our own.

Note: Moon phases are from the historical lunar tables available on astropixels.com (http://astropixels.com/ephemeris/phasescat/phases0701.html).

References Cited

Aulie, William H., and Evelyn W. de Aulie. 1978. Diccionario Ch’ol-Español, Español-Ch’ol. Mexico City: Instituto Lingüistico de Verano.

Grube, Nikolai. 1992. Classic Maya Dance: Evidence from Hieroglyphs and Iconography. Ancient Mesoamerica 3(2):201-218

Hopkins, Nicholas A., Ausencio Cruz Guzmán, and Kathryn Josserand. 2008. A Chol (Mayan) Vocabulary from 1789. International Journal of American Linguistics 74(1).

Houston, Stephen. 2012. Heavenly Bodies. Maya Decipherment, July 16, 2012. https://mayadecipherment.com/2012/07/16/heavenly-bodies/

Houston, Stephen, David Stuart, and Marc Zender. 2017. The Lizard King. Maya Decipherment, June 15, 2017. https://mayadecipherment.com/2017/06/15/the-lizard-king/

Kaufman, Terence, and John Justeson. 2003. A Preliminary Mayan Etymological Dictionary.

Kaufman, Terrence, and William Norman., 1984. An Outline of Proto-Cholan Phonology, Morphology and Vocabulary. In: J.S. Justeson and Lyle Campbell (eds.), Phoneticism in Mayan Hieroglyphic Writing, 77-166. Institute for Mesoamerican Studies Publication No. 9. SUNY, Albany.

Keller, Kathryn, and Plácido Luciano G.. 1997. Diccionario Chontal de Tabasco (Mayense). Serie de Vocabularios y Diccionarios Idígenas “Mariano Silva y Aceves,” Número 36. Tucson: Summer Institute of Linguistics.

Knowles, Susan Marie. 1984. A Descriptive Grammar of Chontal Maya (San Carlos Dialect). Ph.D. Dissertation, Tulane University.

Martin, Simon, and Joel Skidmore. 2012. Exploring the 584286 Correlation between Maya and European Calendars. The PARI Journal 13(2):3-16.

MacLeod, Barbara. 1990. Deciphering the Primary Standard Sequence. Ph.D. Dissertation, Department of Anthropology, University of Texas, Austin, TX.

Morán, Pedro [sic]. 1935. Arte y diccionario en lengua cholti. Edited by William Gates. Maya Society publ. 9.

Polian, Gilles. 2020. Tseltal-Spanish multidialectal dictionary.
Dictionaria 10. 1-8109. DOI: 10.5281/zenodo.3660891 (Available online at https://dictionaria.clld.org/contributions/tseltal, Accessed on 2020-08-01.)

Prager, Christian. 2020. A New Logogram for “to Arrive” – Implications for the Decipherment of the Month Name Cumku. Textdatenbank und Wörterbuch des Klassischen Maya, Research Note 13. Universität Bonn.

Stuart, David. 2012. The Verb Bix “Go, Go Away.” Maya Decipherment, January 23, 2012. https://mayadecipherment.com/2012/01/23/the-verb-bix-go-go-away/

Stuart, David, and Marc Zender. 2018. Epigraphy and History at La Corona: Contingency and Surprise in Epigraphic Discovery.  Paper presented at the 83rd Annual Meetings of the SAA, Washington, D.C.

Toxxer, Alfred M. 1941. Landa’s Relación de las Cosas de Yucatan: A Translation. Papers of the Peabody Museum of American Archaeology and Ethnology, vol. XVIII. Peabody Museum, Cambridge, MA.

Wisdom, Charles. 1950. Materials of the Chorti Language. Middle American Cultural Anthropology Microfilm Series 5, Item 28. University of Chicago Library.