A Possible Sign for Metate 9

by David Stuart, Department of Art and Art History, The University of Texas at Austin

The "Bent-Cauac" Sign

The “Bent-Cauac”

Among the still-undeciphered signs in Maya writing is the so-called  “bent-cauac” element (Figure 1). Most epigrpahers seem to agree that it is a logogram (a word sign), but its precise reading has so far remained elusive. In this note I would like to put forth some evidence that points to a possible reading KA’, with the meaning “metate” or “grinding stone.” The reading, if correct, may ultimately help us to understand a key place name cited in historical records of the Classic period.

Figure 2. The "maguey metate" place name. (a) TIK St 31, (b) COL La Florida(?) vessel, (c) COP St 4, (d) COL vessel K1882. Drawings by D. Stuart; K1882 Photo by J. Kerr.

Figure 2. The “chi-altar” place glyph. (a) TIK St 31, (b) COL La Florida(?) vessel, (c) COP St 4, (d) COL vessel K1882. Drawings by D. Stuart; K1882 photo by J. Kerr.

The bent-cauac sign is perhaps best known as part of an important place name in early Maya history, mentioned in the inscriptions from a number of different sites, including Copan, Tikal and Dzibanche. as well as depicted on a few codex-style ceramics (Grube 2004) (Figure 2). Here it is combined with the hand sign chi, which some years ago led to the nickname “chi-witz” (Grube 2004:127) apparently based on the bent-cauac’s imperfect resemblance to the WITZ, “mountain,” logogram identified a number of years ago (Stuart 1987). Clearly it is a different sign, however.[Note 1] More recently, some epigraphers have opted to refer to the place name as “chi-altar,” seeing a connection instead to the large table-like altars sometimes depicted in Maya sculpture and painting (see for example Stone and Zender 2012:93). This visual connection to a stone object seems closer to the mark, yet I believe the “altar” designation remains vague and even problematic. One reason for my hesitance is the distinctive and consistent bent form of the sign’s main element — something altogether different from the flat altar stones with two supports. Moreover, a hieroglyphic sign that actually does depict such stone altars or tables already exists in the texts of Tikal and Copan. Significantly, one inscription at Tikal includes both the the “bent-cauac” and “stone table” signs, easily demonstrating the distinction of the two elements (See Tikal Stela 26, blocks zA7 and yB2). Thus there is good reason to see the bent-cauac as neither a hill nor an altar, but representing some other type of stone object or feature.

Figure 3. Corn-grinding scene on K1272 (Photograph by J. Kerr).

Figure 3. Corn-grinding scene on K1272 (J. Kerr photo).

If we look at the bent-cauac’s visual history, we see that the sign changes somewhat over time. Its earliest known cases show two small stone elements below the larger bent sign (Figure 2a). Later scribes usually opted to place small stones at the upper left and lower right corners of the sign (Figure 1, Figure 2c, Figure 4), lending the sign  aesthetic and visual balance.  In some instances, the smaller stone elements are omitted altogether (Figure 2b).  In the iconographic parallels from codex-style vases, we see that the original early form is retained, showing an irregular, sloped large stone atop two supports (Figure 2d).

In considering what the bent-cauac sign really depicts, we can be sure of a couple of things. One, it is a stone object of irregular shape, sloping downward on one end. Second, it can have “supports” of stone, but not always. What might it be? I suggest that it probably represents a metate, or a grinding stone — an identification that seems to agree well with the depictions of such objects in Maya art (Figure 3).  In the fuller examples of bent-cauac logogram (see Figure 1), the placement of a stone on top may allude to the hand-held “mano,” with the other stone serving as a support beneath.

FIgure 4. Example of the -a suffix on the bent-cauac

Figure 4. Example of the -a suffix on the bent-cauac

Some phonetic evidence may help determine the sign’s value.  In various instances we see the bent-cauac sign with an -a suffix (Figure 2c, Figure 4). This is a sign that in its origin represented a parrot’s beak, abbreviating the fuller parrot head sign also a, also seen conflated with the metate glyph in cases from the Hieroglyphic Stairway at the site of Resbalon. In this context the –a suffix sign can be taken in a couple of ways. The –a element might conceivably be providing the common place name ending  –(h)a’, “water,” as it clearly seems to do in the Yaxha toponym and emblem glyph (YAX-a) (Stuart 1988). Alternatively, the –a may provide a telling phonetic clue to the reading of the logogram, serving as a phonetic complement.

I prefer this second possibility, since it seems to be an optional sign added onto the metate sign in at least two separate contexts.  If the -a is indeed optional, there is a good likelihood that it serves a phonetic complement to the reading of the metate logogram.  In this light, it is interesting to see the various terms for metate in lowland Mayan languages, as listed by Kaufman in his Preliminary Mayan Etymological Dictionary (Kaufman 2003).  There the form reconstructed for proto-Mayan is *ka’, and for Proto-Ch’olan it is *cha’.  I therefore suggest this may be a good working decipherment for the bent cauac sign, either KA’ or CHA’, “metate.”

Metates were, of course, basic implements in domestic food production used throughout the ancient Americas. In Mesoamerica we usually think of stone grinders being used for processing maize, but they were key implements in many different types of food preparation. Interestingly, metates were used for the grinding of maguey and other agave plants in the manufacture of mescal, pulque and perhaps other fermented drinks important in Mesoamerica.  

We might now have a reasonable interpretation of the mysterious place glyph once called “chi-witz.” If I were to propose a phonetic analysis of the compound, something like chi-CHA’ (chi(h) cha’) or chi-KA’ (chi(h) ka’), the “maguey grinder (place),” looks like a workable possibility.

It is important to stress that the geographic frame of reference for this “maguey-grinder” place name still remains very unclear. Some have argued that it might refer to El Mirador or Nakbe, given its early historical connections (see Grube 2004:13-131; Zender and Stone 2012:234). While such connections are tantalizing they still remain circumstantial, and without further evidence it is difficult to know. Perhaps this better semantic understanding of the place name will help us one day in resolving the issue.

It is also important to note that not all appearances of the supposed metate sign are easily understood, even if KA’ or CHA’ turns out to be a correct reading. On Tikal Stela 26 the sign appears in what might be a verbal context (U-KA’-ji) but the surrounding text is obscure. Hopefully these and other issues can be clarified with further analysis.


Thanks to Stephen Houston, Simon Martin and Karl Taube for some very useful feedback on this proposal.


Note 1. Part of the confusion seems to have stemmed from an example from Stela 1 at Arroyo de Piedra (see Grube 2004:130), where the sculptor of the monument bears the title CHIH-WITZ AJAW, “Deer-Mountain Lord.” There is no reason to connect this isolated example of the “Deer Mountain” place name to the “chi-witz” or “chi-altar” glyphs under discussion here, however.

Note 2. The difference in these two readings rests on whether one prefers to transcribe the sign using the reconstructed Ch’olan-Tzeltalan form cha’, or the more “archaic” ka’.  Until recently I would have opted strongly for the latter, given the secure position of Classic Mayan language in the Ch’olan-Tzeltalan group. But it is important to point out that many glyphic spellings point to a more complex scenario of areal diffusion of the k > ch sound change, and that the supposed innovation is not as regular as was earlier assumed (Law, et. al., in press). Until further clarification comes about, KA’ or CHA’ seem equally plausible readings.


Grube, Nikolai. 2004. El origin de la dinastia Kaan. In Los Cautivos de Dzibanche, edited by Enrique Nalda, pp. 117-132. INAH, Mexico D.F.

Kaufman, Terrence. 2003. Preliminary Mayan Etymological Dictionary. On-line resource at http://www.famsi.org/reports/01051/pmed.pdf.

Law, Danny, John Robertson, Stephen Houston, Marc Zender, David Stuart. In press. Areal Shifts in Maya Phonology. Ms. accepted for publication in Ancient Mesoamerica.

Stone, Andrea, and Marc Zender. 2012. Reading Maya Art: A Hieroglyphic Guide to Ancient Maya Painting and Sculpture. Thames and Hudson, New York.

Stuart, David. 1985. The Yaxha Emblem Glyph as YAX-A. Research Reports on Ancient Maya Writing, 1. The Center for Maya Research, Washington, D.C.

_________. 1987. Ten Phonetic Syllables. Research Reports on Ancient Maya Writing, 14. The Center for Maya Research, Washington, D.C.


  1. Let me throw out an idea on the reading of Tikal Stela 26 concerning the “Matate” place in the context of other place names on the stela.

    The primary text can be read as a list of three objects/places, each “owned” by three supernaturals (not to be confused with the rulers who adopted their names).

    The first in the list refers to the stela itself, owned by Chanal Chak Bay Kan, Sihyaj Chan K’awiil, and Chak Ich’aak. (B2-B4)

    The third in the list is the Matate Place that you describe, owned by Baby Jaguar, Itzamnaaj, and Ehb Xook. (A7-A9)

    The second in the list is less obvious but glyph block A5 is the key. The glyphs appear to read u-K’AK-BALAM-na-li > u k’ahk bahlamnil > “their K’ahk’ Bahlam house.” Next come the names of the three gods who own this structure (B5-B6).

    (In this reading I am assuming an early form of the syllable –na- that Mora-Marin has documented, and that the -na-li suffixes denote a house/structure/compound that is possessed, used just as they are in u-pi-bi-na-li > “his sweatbath.”)

    If this reading is correct, it mentions a place name at Tikal that as far as I can tell has not been previously noted. Its name would be k’ahk bahlamnil > “Fire Jaguar House.”

    It also presents a very “poetic” and symmetric statement of sacred objects and places. If the second place is the sacred compound in which the stela stood, then the description starts with the stela itself, moves out to the sacred area, and then moves out to the wider city.

    Given the sacred nature of the other two items in the list, Stela 26 would seem to imply that the Matate Place is as much a sacred center as a secular city.

    (Mark Pitts)

  2. Thanks for your comment, Mark. The inscription on Stela 26 is very opaque, though I hesitate to see the text on the left side as naming a series of sequential places or objects. Not that I have any great alternatives to offer, mind you. The text seems to link the stela to the current ruler K’an Kitam and the ancestral figures he conjures (his father Siyaj Kaan K’awiil and earlier predecessor Chak Tok Ich’aak) on the unknown Period Ending being commemorated. What’s dicey here is the glyph you mention at position zA5 (using the Tikal Report designation). It is clearly a possessed noun, but I see nothing there about houses or architecture (i.e., no NAAH sign). Instead the element below the jaguar head looks to be the element I identified some years ago as the “blood” logogram, probably read K’IK’. I’d hate to hazard a reading for this whole combination, as it’s a term I’ve never seen anywhere else. Are we looking at a reference to jaguar blood and sacrifice, a la Altar Q at Copan? And as for the “metate” reference a bit later on, I don’t see how it can be a place name in this setting. It lacks the maguey component, and instead has a -ji suffix, suggesting a very different role. I agree that we are looking at some unusually vivid poetic language on this monument, and I’ve lately wondered if the possible metate glyph (U-KA’?-ji) might somehow connect with the growing maize plant depicted in the oddly written TZ’AP verb above (zA2). Is the text linking the “planting” of the stela to a subsequent “grinding” of offerings for the deity Triad? Maybe a metaphoric reference to the stela as maize-stalk, and the altar as metate?? Without the rest of this gorgeous text I don’t think we can ever know. All very esoteric stuff.

  3. David, Thanks for the feedback. As followup, I’m wondering about the 3 triadic groups of beings (as opposed to the nouns they possess). If some are humans and some are gods, would this be fairly unique? – that is, having gods and humans mixed together in triadic groups?

  4. Un comentario con respecto a la nota 2. Son comunes los procesos de palatalización de ciertas consonantes (en este caso una velar /k/) ante vocales anteriores (en este caso /i/ y /e/) dando como resultado una africada /ch/. Lo más probable es que siempre encuentren la /ch/ ante estas dos vocales (como en chitam, chih) cuando en otras lenguas se tiene /k/. Lo que me llama la atención, y me agrada, es la variación /k/-/ch/ ante la vocal /a/ (no en la misma palabra). Es interesante porque me da la idea de que habría que reconsiderar la propuesta de la existencia de una serie de consonantes velares palatalizadas /ky/ y /ky’/ en protomaya. El Mam tiene estas consonantes y se torna más interesante cuando miramos esos mismos cognados a nivel dialectal en esa lengua pues en Ixtahuacán dicen, por ejemplo, kyaj ‘cuatro’, kya’j ‘cielo’, kyaq ‘rojo’ y kyeej ‘venado’, mientras que en Todos Santos dicen chaj, cha’j, chaq y cheej respectivamente. En tz’tujil, si bien no se describe que el sistema consonántico tenga /ky/ y /ky’/ las presenta fonéticamente en las palabras kyeej ‘caballo’, kyaq ‘rojo’ y ky’aq ‘pulga’ (Pérez y Hernández 1996).
    Yo tengo la hipótesis de que sí habían /ky/ y /ky’/ en protomaya y que en lenguas cholano-tzeltalanas se volvieron /ch/ y /ch’/ ante cualquier consonante (en un estado muy temprano) y, que después del cambio de /q/ y /q’/ a /k/ y /k’/ en cholano, tzeltalano y yucatecano, lo que originalmente era /k/ y /k’/ en cholano-tzeltalano se acomodó en /ch/ y /ch’/ en una etapa posterior.
    El ejemplo de ‘pulga’ es interesante (en Mam: ky’aq) porque si vemos en lenguas yucatecas ‘pulga’ se dice ch’ik mientras que en tzeltal, por ejemplo, es ch’ak. La reconstrucción de Kaufman (2003) es *k’aq y la idea es que la correspondencia esperada en lenguas yucatecanas sería *k’iq. Las lenguas yucatecas son para mí la clave porque es posible ver que las consonantes *ty y *ty’ del protomaya se palatalizan en /ch/ y /ch’/ en las lenguas yucatecas en dos contextos: ante /i/ y /e/, ej. *tyee’ => che’, *tyii’ =>chi’; y también a final de palabra, ej. *paaty => paach ‘espalda’ (en tzeltal pat). Al parecer, con el ejemplo de ‘pulga’ vemos que, lo que es seguro, es que hubo un cambio de a => i, y sólo me explico esa palatalización por dos motivos: 1. que sea un préstamo (dudo que ‘pulga’ sea un préstamo, aunque sí es más probable que chak ‘rojo’ lo sea), 2. que haya tenido una /ky’/ que se palataliza ante /i/ como lo hicieron la /ty/ y la /ty’/ en estos contextos. Lo mismo puede decirse de la palabra ‘pozo’ en yucateco, ch’e’en, la reconstrucción de Kaufman (2003) es *k’e’n, no es la correspondencia esperada, y puede explicarse por las mismas razones anteriores. Yo creo que el Mam es clave para esto, y en esa lengua ‘metate’ se dice ka’, no kya’, mientras que ‘panela’ es kab. En los códices veo que ‘abeja’ (o ‘miel’) está como KAB. De acuerdo con mi hipótesis la lectura (para en verdad una forma muy arcaica) sería KA’. Existe otra posibilidad y es lo que de acuerdo con England (1983) ocurre en mam, ella dice que /ky/ y /ky’/ son innovaciones del mam que aún se encuentran en distribución complementaria parcial con /k/ y /k’/ pero describe un contexto en el que las primeras aparecen y es ante /i/, /e/ y /a/ cuando están seguidas de una oclusiva uvular o fricativa velar, el mismo contexto mencionado en Pérez y Hernández (1996), aunque sólo fonéticamente, para el tz’utujil. Es posible que en maya clásico “más arcaico” (una forma restituida) existiera este mismo proceso de palatalización en este contexto dando como resultado /ch/ y /ch’/, en una etapa posterior en palabras como ‘rojo’, mientras todavía se tiene /k/ y /K’/ en otras como en ‘culebra’. El ejemplo de ‘rojo’ es interesante porque en lenguas tzeltanas vemos que es tsaj/tsoj, con una africada alveolar, similar al proceso de *ty que en mam se vuelve /ts/.Creo que la epigrafía maya ayudará mucho a resolver cuestiones históricas de la familia lingüística.

    Veo que en yucateco y chorti’ ji’ significa ‘arena’ como sustantivo. Y como verbo en yucateco es ‘alisar’, y en chorti’ es la raíz verbal de ‘descomponer, deshacer piedras chiquitas. ¿Qué posibilidades hay de U-KA’-ji pueda ser u ka’ ji’ ‘la piedra de/para moler’?

    Saludos cordiales

    Martín Sobrino

  5. Hello David – I always wondered about the prefixed syllable /ko/. Here are some speculations. Several instances are known where T316 is prefixed by T110 /ko/, such as on Kerr 2573, NAR Alt. 1 or the CRC Stuccoes. This is of course not a new observation but is contained in several publications and workshop books of the past ten or fifteen years or so. I know that the prefixed /ko/ has elsewhere been suggested to express /kok/ and /kok witz/, but this syllable with the initial /k-/ is one of the rare prefixed complements indicating a possible K- value to T316. Prefixed phonetic complements of the group k- indicate a root that is “affected by the consonant shift from proto-Mayan *k > proto-Cholan-Tzeltalan *ch” according to a 2005 paper by Nikolai. If /ko/ is indeed a phonetic complement and not used as logograph /KOK/ then we have to consider that T316 is either CH- or K-, maybe ending in -ji, koj or choj?? Interestingly on NAR St. 35, a war-related monument with references to mythological events, the expression “T316-li”, introduced by /ha-i/, is the object of the verb /u-ko-bo/ > u-kob “to create” after a passage mentioning the supernaturals CHAN-TE’ XIB, CHAN-TE’ CH’OK. According to the narrative they burned the JGU, the visual has been summarized by Marc in his Reading Maya Art book, p. 32. According to Naranjo Stela 35 these agents seem to “kob” or create / engendered the T316-object after the burning of the jaguar-baby at the beginning of the sequence. Maybe, after the burning of the JGU, CHAN-TE’ XIB and CHAN-TE’ CH’OK engendered the T316-object by putting something new on it?

    Best, Christian Prager

    • Hi Christian, Thanks for the interesting observations. I would agree that the phonetic evidence has its ambiguities, and, as I have long argued, a proposed reading of a sign can’t rely solely on an understanding of what it represents, i.e. its iconicity (to use academic jargon). Having said that, I’m fairly convinced that the sign represents a metate, even if the actual reading of the sign might prove to be something different from what I proposed. No doubt the ko- prefixes are an important part of this puzzle, but I’m not sure how they are working in these instances. In the Caracol B16 stuccoes, the ko-“bent cauac” glyph really looks to be a place name, maybe for some location near Naranjo, or at least affiliated with it. If we take ko- to be a phonetic complement in these instances (and I agree it’s not for kok), then we’d have to analyze the place name as nothing more than the reading of the “bent-cauac” itself, whatever that might be. This isn’t outside the realm of possibility, but it seems somewhat unlikely to me, since most toponyms (as opposed to emblems) are composed of more than a single lexeme. So I wonder if the “ko” has some other role here that we don’t understand, and that it isn’t serving as a straight complement to the reading of the bent-cauac. It all makes for a curious pattern, so I’m not sure any of this is satisfactory. I’ll keep my mind open, though, and see where it leads. By the way, I doubt the -ji suffix on Tikal Stela 26 is a phonetic complement, either. There I suspect it’s providing a morphemic suffix after the logogram.

  6. Hi David. I just noticed that another context of T671:316 is found on IKL Lnt. 1 (unfortunately I cannot include a picture). It is contained in the nominal phrase of the female owning the waybil. Interestingly the name phrase reads T1000a.[.]:671:316:126 and the question is now whether T671:316 is a complex sign representing the hand on the metate or the chi-throne, as suggested previously. Most importantly, the suffix after this sign is /ya/. Is it a complement or a nominalization suffix /-ya/, which seems not very probable since T671:316 is seemingly not a verb.

  7. Dear Professor Stuart,

    It is just one thought regarding the collocation ko-KA’/CHA’. In the Maya Dictionary by Barrera Vasquez, we find the translation of the word “ko” as “grano de algo, grano, grano de maiz”. We wonder if it could work in this case because “grains grinder / metate para granos” makes a sense.

    Kind regards

    Boguchwala Tuszynska and Kajetan Jagodzinski

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s