The earliest known example of Maya writing has been documented at the site of San Bartolo, Guatemala (map), as reported in an article by William A. Saturno, David Stuart, and Boris Beltrán (2006) on the website of the journal Science.

The article, entitled "Early Maya Writing at San Bartolo, Guatemala," may be downloaded from the website of the San Bartolo Project.

Perhaps most intriguing in the Science report is the indication that Maya writing might have influenced the so-called Isthmian script, rather than the other way around (ibid.:2). Even more provocative is the consideration that the writing from San Bartolo is not only the earliest known for the Maya, but verging in its dating on being among the earliest scripts from anywhere in Mesoamerica (ibid.:1).

It has been asserted that the earliest writing in Mesoamerica appears on an Olmec cylinder seal from the vicinity of La Venta, Tabasco, dating to about 650 BC (see Mesoweb report). But the supposed glyphs are more likely to be iconic elements than signs that form a continuous text (Houston 2004b:203.) True writing is a graphical representation of spoken language, and it is "disposed into linear sequences that can theoretically expand into greater degrees of syntactic complexity" (ibid.:292). By contrast, iconic elements—or codified symbols—can be joined in meaningful arrangements but do not clearly record the sounds of language. A knowledgeable viewer could attach a word to a given element, but there is no fixed sequence or reading order; only the overall arrangement—or "emblem"— makes sense of the individual components (ibid.:284, 286).

For instance, the graphemes on a greenstone tablet from Guerrero, Mexico, do not constitute writing but rather symbols representing a hill (a recurrent element in Mesoamerican toponyms), beneath a directional symbol with vegetal elements, and a headdress (which might designate a lordly title) (ibid.:286).

A cylindrical ceramic roller stamp from Tlatilco has also been adduced as writing (Flannery et al. 2005:Fig. 3) but actually "shows a well-known icon of centrality, a flower, and a deity head" (Houston 2004b:288). Given that these cylinders were apparently used to roll out a repeating pattern on human skin or textile, it would have been impossible to say where the supposed text began (ibid.:288).

The first true writing in Mesoamerica appears either on La Venta Monument 13 or Monument 3 of San José Mogote (ibid.:292). La Venta Monument 13, the so-called "Ambassador" (Drucker, Heizer, and Squier 1959:40), displays four glyphs, three in a linear arrangement and the fourth a footprint, the common Mesoamerican symbol for movement. The carved stone monument may date to 600-400 BC (Houston 2004b:292), although Karl Taube considers a dating as many as two hundred years later (personal communication 2002, cited in Houston 2004b:276, 292).

The San José Mogote monument depicts a nude man, presumably a captive, with two glyphs between his feet that have been interpreted as a day name in the Zapotec 260-day calendar (Marcus 1992:36). Later captive monuments from Oaxaca are marked by true linear texts (e.g. Monte Albán Danzante MA-D-55 [Houston 2004b:Fig. 10.1a]), so the glyphs of the San José Mogote monument are reasonably construed as elements of writing. But opinions differ as to the date of Monument 3, ranging from c. 600 BC (Marcus 1992:36) to as late as 200 BC or even later (Cahn and Winter 1993, cited in Houston 2004b:293).

Writing that probably encodes the Zapotec language appears in Oaxaca as early as 500 BC, most familiarly on the so-called Danzantes of Monte Albán, "disemboweled, castrated, and bloodied captives in various states of agony or post-mortem repose" (Houston 2004b:293).

There are only a very limited number of early Zapotec inscriptions, while examples of the later Isthmian script are even fewer (ibid.:293, 296). Named for the region from which they derive, in and around the Isthmus of Tehuantepec, "Isthmian" is intended as a neutral term for these inscriptions, as distinct from "Epi-Olmec," with its unproven assumption, based on linguistic geography and the regional study of loanwords, of an Olmec derivation (ibid.:296). An inscribed mask from a private collection has recently expanded the corpus of Isthmian signs (see Mesoweb report).

There are very few examples of Maya writing from the Late Preclassic period (400 BC to AD 200), and most are from looted objects that can only be dated on the basis of style, probably to no earlier than c. 100 BC - AD 100 (Saturno, Stuart, and Beltrán (2006:1, 2). Of the few stone monuments, El Mirador Stela 2 (Hansen 1991) dates in all likelihood to no earlier than 100 BC, again on stylistic grounds. A monument with glyphs excavated at El Portón, Guatemala (Sharer and Sedat 1987:49-73, Plate 3.8; Sharer and Traxler 2006:Fig. 5.9), may be dated to as early as 300 BC, but only on the basis of a solitary radiocarbon date not in direct association with the carving (Saturno, Stuart, and Beltrán 2006:2).

The Preclassic murals of San Bartolo were discovered in 2001, when Willliam Saturno, exhausted and on the verge of dehydration, sought shade in a looters' tunnel (Saturno, Taube, and Stuart 2005:2). Excavations of the San Bartolo Structure 1—also known as Las Pinturas for the paintings within—may have begun illicitly, but they were continued scientifically by El Proyecto San Bartolo, beginning in March 2003. The environment was stabilized and monitored, the paintings exposed and cleaned without chemicals, then completely documented by photography, flatbed scanning, and on-site drawings by professional artists, iconographers, and epigraphers (Saturno, Taube, and Stuart 2005:6; Saturno 2006:76-77; also see video by David Pentecost at The Daily Glyph).

The north wall of the murals chamber was the first to be exposed (Saturno, Taube, and Stuart 2005). The central character in its colorful scene is the Maya maize god, whose features, not at all coincidentally, are distinctly Olmec (ibid.:25-28). When the west wall was subsequently exposed, its complex mythological scene also featured the maize god (Saturno 2006; see Mesoweb report).

Painted glyphs in the murals room are some of the earliest securely dated Maya writing, as radiocarbon analysis dates the room itself to about 100 BC (Saturno, Stuart, and Beltrán 2006:1). Given the early date and the dearth of inscriptions for comparative analysis, only the syllables po, mo, ja, and the logogram AJAW, "lord, ruler," can be deciphered with any confidence (Saturno, Taube, and Stuart 2005:41-48). As team epigrapher David Stuart observes:

The opaque nature of these texts seems best explained by their extremely early date... The hieroglyphs so far exposed in the San Bartolo murals are among the earliest examples of lowland Maya writing, and there are few other Preclassic texts from lowland sites with which we can compare them. A near-contemporary inscription from Stela 10 at Kaminaljuyú, in the highlands, looks significantly different from the form of writing evident at San Bartolo, and we are willing to posit that the writing systems represented at these two sites are quite different. It has been widely assumed that the writing of Kaminaljuyú was the direct precursor of lowland Maya writing, but the initial evidence from San Bartolo raises the possibility that Maya writing in the lowlands had a far older and independent history. (ibid.:46-47)

The Pinturas pyramid was amassed in seven construction episodes, with the structures atop previous phases either partially or totally destroyed before being built over in Precolumbian Maya fashion. The murals chamber owes its preservation to the fact that it was filled in and covered over. As the penultimate phase, it is designated Sub-I.

It was explorations in 2005 of a phase four construction episodes earlier that resulted in the discovery, by Boris Beltrán, of a stone block from a dismantled wall of the structure that had once stood atop the Sub-V platform (Saturno, Stuart, and Beltrán 2006:1). The ten glyphs that presently constitute the earliest known Maya writing were painted on this block in a calligraphic hand. The bygone central chamber of Sub-V was decorated with colorful murals; the doorjamb still displays a polychrome image of the maize god, the central character from the previously discovered mythological scenes (ibid.:1). The column of glyphs, apparently part of a longer sequence, may have been associated with this maize god mythology, but the dismantling of the wall for the subsequent construction phase makes it impossible to say where the glyph-painted block was located in the room (ibid.:1).

The text has been dated to as early as 300 BC and no later than 200 BC on the basis of accelerator mass spectrometry radiocarbon dates of five charcoal samples from sealed contexts in Sub-V and the two construction phases that bracket it (ibid.:1). One from within the floor of Sub-VI establishes the terminus post quem (the earliest possible date for the room and its text); the other four are from within the floor of Sub-V and from the fill surrounding the painted blocks, relating therefore to the destruction of the Sub-V painted room and the construction of the Sub-IV platform above it. Taken together with samples from the final two construction phases, they suggest that the glyphs were painted between 300 and 200 BC (ibid.:1).

Again the only fully decipherable glyph is AJAW. In overall appearance, there is some resemblance to the Isthmian script. "All examples of that script post-date the San Bartolo block, however, raising the question of what direction any influence may have flowed" (ibid.:2).

Saturno, Stuart, and Beltrán raise the question of the relationship of Maya writing to early writing elsewhere in Mesoamerica, given the early date of the San Bartolo glyphs. Noting that Preclassic writing systems were clearly established by about 400 BC in Oaxaca and possibly the Isthmus of Tehuantepec, they conclude (ibid.:2): "It now appears that the Maya also participated in the Pre-classic cultures of literacy, and at a significantly earlier date than previously believed."