Updated April 2005
In a report published December 6, 2002, in the journal Science (Vol. 298, p. 1984), archaeologists led by Dr. Mary E. D. Pohl of Florida State University in Tallahassee announced the discovery of what they describe as writing symbols, or glyphs, on a cylinder seal used for printing and also on fragments of a greenstone plaque. These artifacts were excavated near the Olmec site of La Venta, Tabasco, Mexico and dated to about 650 BC.
A webpage from Dr. Pohl's department at Florida State University provides a short discussion the claim to have found the earliest writing as well as photos of the cylinder and two greenstone fragments which are claimed to have logographic writing on them. There is also a link to a draft of the Science article on the find (download).
A report on the archaeological investigation in which the find was made is now available at FAMSI: Olmec Civilization at San Andrés, Tabasco, México, by Mary Pohl, with contributions by Christopher von Nagy, Allison Perrett and Kevin Pope.
Dr. Pohl also describes the find in a National Public Radio broadcast.
In a discussion of early Mesoamerican writing systems that does not necessarily bear directly on the find by Pohl and her colleagues, Stephen Houston (2004b) offers observations that may be tangentially applicable. Speaking of the system of iconography or "codified symbols" that has been labeled Olmec, Houston explains that these clearly derive from nature or from artifacts; for example, a hand is depicted apart from the rest of the body (Houston 2004b:284, citing Justeson and Mathews 1990:90-93). The Olmec system employs the principle of pars pro toto, whereby one part represents or stands for the whole.
This enables a certain economy of representation and permits the clustering of elements that imply more complex, if implicit, orderings. But it would be incorrect to see such clusterings as writing per se which, following most authors in this volume, can be defined as graphs that record elements of language. In Olmec iconography, the juxtaposition of elements refers to spatial organization, place names, and icons of centrality, including directional symbolism: the elements exist, not as text, but as features of a landscape, including primordial, cosmic ones (Reilly 1995:38-39). In this sense, these images or graphs might be best understood as emblems, which join elements into meaningful arrangements yet do not clearly record sound. (Houston 2004b:284)
Referring to the individual components of these emblems, Houston continues:
A knowledgeable viewer could look at these elements and attach a word to them. However, that procedure of lexical identification would be irrelevant to apprehending the full meaning of the emblem. Places and titles linked to them could be understood in any language, a benefit of an "open," inclusive system of communication that bridged many different groups. Yet the elements are not sequenced. They have no fixed reading order. Only the overall emblem, each part taking meaning from the whole, makes sense of its constituents. As a consequence, if writing is the recording and ordering of lexemes, then this is not writing. (Houston 2004b:285)
With regard to stone or clay cylinders, Houston observes that surviving images of Olmec bodies bear designs that might have been produced by rolling such cylinders across the skin, although they might also have been produced by tattooing (2004b:287).
Virtually all the designs that can easily be interpreted are purely iconographic, and the lone example thought to be a text, from the highland Mexican site of Tlatilco, shows a well-known icon of centrality, a flower, and a deity head (Reilly 1994:242-243). Repetition of the design as it rolled out on the receiving surface would have made it difficult to determine where the "text" began. (Houston 2004b:288)
Houston does speak of the La Venta find in particular:
This section cannot end without addressing recent claims for "Olmec writing" at about 650 BC in La Venta, Tabasco (Pohl, Pope, and von Nagy 2002). The finds consist of two incised greenstone objects of small size. The "signs" are rounded, as Mesoamerican hieroglyphs tend to be, but do not appear to form continuous text or correspond to elements from later scripts, contrary to claims by the authors. The more cautious view of these two objects not evident in the massive publicity generated by journalists (e.g. www.newscientist.com/news/news.jsp?id=ns99993151) is that the finds are suggestive yet inconclusive. The stronger likelihood is that they are iconic elements. (Houston 2004b:293)
In early January 2003, Erik Boot contributed further links and information, as follows:
NewScientist.com indicates that the drawing on the BBC site was made by Ayax Moreno. The article itself is short (an online version of the article published in Science).