CELTIFORM STELAE
A NEW OLMEC SCULPTURE TYPE AND ITS IMPLICATIONS FOR EPIGRAPHERS
James B. Porter
December 25, 1989
INTRODUCTION TRAITS SUBTYPES CONCLUSION
During July of 1984 I visited Rebecca Gonzalez Lauck's survey project at La Venta and was shown a portion of La Venta Monument 27 which local inhabitants had partially exposed in the belief that it was the door to a secret treasure chamber within the earthen heart of the site's great pyramid. From La Venta I proceeded to the Tres Zapotes site museum and subsequently to the national museum in Mexico City, to research corrections of my preliminary drawing of the two extant fragments of Tres Zapotes Stela C which are preserved in those two INAH facilities. Upon returning to Berkeley to correct my drawing of Tres Zapotes Stela C, I recalled the mythic purport of Monument 27 and decided to examine the photographs of the sculpture in Plate 54 of Drucker et alii (1959). Despite the poor preservation of Monument 27, Plate 54 showed a complete stela with formal and compositional features related to those of Tres Zapotes Stela C, on which I had just been working.
A search of the photographic record of Olmec sculpture revealed that six additional stela fragments, Tzutzuculi Stela 1, Tzutzuculi Monument 1, La Venta Monuments 25, 26, 58 and 66 shared formal relationships with La Venta Monument 27 and Tres Zapotes Stela C. Tzutzuculi Stela 1 was first recorded by Navarrete (1959), Tzutzuculi Monument 1 was discovered by Lowe and McDonald (McDonald 1977), La Venta Monuments 25 and 26 were discovered alongside Monument 27 on the last day of the 1957 field season and were reburied that same day (Drucker et alii 1959:120, 209). La Venta Monuments 58 and 66 were recorded during another field season (Clewlow and Corson 1968). La Venta Monuments 25, 26 and 27 were re-excavated by Gonzalez Lauck (1988:145). Tres Zapotes Stela C had been discovered in two fragments, the first, middle, fragment was found in 1939 (Stirling 1939) and the second, top, fragment was found in 1969 (Beverido 1971, 1987).
Unfortunately, none of the eight Celtiform Stela fragments were found in primary archaeological deposits and the records of their discovery are not particularly useful for assessing their chronological relationships. Navarrete abandoned his original (1959:6) date for the final placement of Tzutzuculi Stela 1 twenty years after his work at the site (Milbrath 1979:27-28). The crude petroglyphs on the back of Tzutzuculi Monument 1 indicate it was originally carved as a freestanding sculpture, only to be reused as one of two monuments flanking the stairway of Mound 4, radiocarbon dated at 660-410 BC (McDonald 1977:561, 1983). Drucker et alii (1959) gave no chronological information for Monuments 25/26 and 27, beyond the C 14 date of 900-200 BC for the platform on which they were re-erected. Clewlow and Corson (1968:178, 180) gave no information on the provenance of La Venta monuments 58 and 66. Unfortunately, C14 age measurements cannot date the carving of stone sculpture and only Tres Zapotes Stela C, with its Maya style hieroglyphic date of 31 BC, is securely dated.
Before beginning a style sequence of Celtiform Stelae, visual records of all eight fragments were collected. During this process I found that Monument 26 bore no "jaguar mask", as described by Drucker et alii (1959:206), nor any "small scale figures", as described by Milbrath (1979:39). However, the search for these features led to the discovery that when Monument 26 was righted (it had been erected upside-down in antiquity, like Monument 27), the break at its lower edge fit the break at the upper edge of Monument 25. Correlations of the dimensions, descriptions and stone of these two sculpture fragments in Drucker et alii (1959:204-206) all indicated that Monuments 25 and 26 were the lower and upper portions of a single original stela. This original stela was designated La Venta Monument 25/26, while the lower and upper portions retain the respective monument numbers 25 and 26 when referenced as separate, reused monuments.
Trait 1, all seven examples are made from unprepared or minimally prepared natural stone slabs.
Trait 2, five examples are divided by a horizontal line incised below the face carved on the front. The two examples from Tzutzuculi treat this division between the face carved on the front and the lower portion of the front differently.
Trait 3, six examples bear a face with buccal mask in front view above the horizontal line. The Mask on Tutzuculi Monument 1 has an unusual type of buccal Mask and La Venta Monument 66 has lost the surface on which this feature was probably carved (Clewlow and Corson 1968:180).
Trait 4, six examples were made from roughly trapezoidal slabs which taper from top to bottom. La Venta Monument 27 does not taper from top to bottom.
Trait 5, five examples bear cheekbands flanking the masked face on either side. Tzutzuculi Stela 1 and La Venta Monument 58 were carved without cheekbands.
Trait 6, five examples bear roughly rectilinear ear ornaments. Tzutzuculi Stela 1 and La Venta Monument 58 were carved without ear ornaments.
Trait 7, four examples bear a basal motif below the horizontal line. Tzutzuculi Monument 1, Tzutzuculi Stela 1 and La Venta Monument 58 were carved without a basal motif.
Trait 8, five examples bear a third sculpture zone above the eyes. La Venta Monuments 58 and 66, have lost their upper portions and cannot bear testimony to this feature although it was probably present on both.
Trait 9, four examples retain a tabbed, or "U", element on the center line in the zone above the eyes. Tzutzuculi Monument 1 was carved without a tabbed element while La Venta Monuments 58 and 66 have lost their upper portions and cannot bear testimony to this feature.
Trait 10, four examples bear a headband with central medallion and elements at the temples in the zone above the eyes. Tzutzuculi Stela 1 was carved without a headband while La Venta Monuments 58 and 66, have lost their upper portions and cannot bear testimony to this feature.
Trait 11, five examples bear conventionalized hair at the top of the carving in the zone above the eyes. On Tuztuculi Monument 1 the hair may be represented by the two curved elements at the upper edge of the carving, on Tzutzuculi Stela 1 the hair is represented by a stepped pyramid element in the zone above the eyes, on La Venta Monuments 25/26 and 27 the hair is represented by scalloped elements at the upper edge of the carving and on Tres Zapotes Stela C the hair is represented by sweeping lines which reach to the upper edge of the carving. La Venta Monuments 58 and 66, have lost their upper portions and cannot bear testimony to this feature.
In addition to variations in specific traits such as basal ornament, variations in surface treatment suggest a general stylistic progression from the simple lines and petroglyphic forms of Tzutzuculi Stela 1 and Tzutzuculi Monument 1 through the increasingly complex lines and forms of La Venta Monuments 58 to La Venta Monuments 25/26, 27 and 66 and culminating in the elaborately detailed lines and forms of Tres Zapotes Stela C. The pecked lines of Tzutzuculi Monument 1 certainly represent an earlier phase of sculptural development than the higher relief of La Venta Monuments 25/26, 58 and 66, while it is possible that the shallower relief of La Venta Monument 27 and Tres Zapotes Stela C represents later sculptural developments. These factors indicate that, aside from variations which might reflect differential artistic development at Tzutzuculi, La Venta and Tres Zapotes, a considerable period of time must have elapsed between the successive carving of the Tzutzuculi, La Venta and Tres Zapotes Celtiform Stelae.
The absence of hieroglyphic texts on the Tzutzuculi and La Venta Celtiform Stelae has serious implications for the popular archaeological belief that the Olmec practiced literacy for some period of time before passing this achievement on to the Maya (Justeson and Kaufman 1993, Stuart 1993a,b). The concrete evidence bearing on this issue, the Maya style hieroglyphic date on Tres Zapotes Stela C and the inscriptions carved on Early Maya Style sculptures, have been a problem for both Olmec and Maya chronology since their discovery (eg. Stirling 1939, Thompson 1943).
Some authors have difficulty accepting the possibility that Olmec style survived until the sixteenth K'atun of Cycle 7 and suggest that the "late" date of Tres Zapotes Stela C is only apparent because the Maya style date on the back was carved by a different people sometime after the Olmec style celtiform theme was carved on the front (Coe 1957). Nonetheless, the chance that the inscription is significantly later than the carving on the front is almost nil. Both fragments of Tres Zapotes Stela C have been examined with precisely this issue in mind and the carving is virtually identical on both sides (Porter 1989a). The forms are defined by shallow areas of relief with scratchy fine line incising and hollow-drilled dots. The method of creating the dots is particularly usefull as a temporal indicator because it is confined to Tres Zapotes Stela C and to Chiapa de Corzo Stela 2, both bearing inscriptions in the sixteenth K'atun of Cycle 7. Also, and perhaps most telling, dots on both the front and the back of Stela C are hollow-drilled to the same circumference and may even have been made with the same (cane?) bit!
Other authors have difficulty accepting the possibility that Early Maya Style sculptures bore inscriptions prior to the sixteenth K'atun of Cycle 7 and simply ignore published suggestions that inscriptions on Early Maya Style sculptures may be Early Maya inscriptions (Justeson and Kaufman 1993, Stuart 1993a,b). Nonetheless, Early Maya Style sculptures bearing Early Maya Style inscriptions probably dating prior to the sixteenth K'atun of Cycle 7 have been identified in several important publications (Heath-Jones 1959, [1970], Miles 1965, Graham, et al 1978, Graham and Porter 1989, Hanson 1991).
Prior to the recognition of antecedents to Stela C it was possible to believe that the absence of earlier Olmec texts was only apparent and that, when earlier examples of Stela C's monument type were found, they would also bear (earlier) "Olmec writing". However, there is little doubt that the Celtiform Stelae of La Venta are earlier than Stela C (Coe 1957:599) and the stelae of Tzutzuculi are earlier still. Nor can it be doubted that these carvings are the long sought Olmec antecedents to Stela C. Also, the Tzutzuculi, La Venta and Tres Zapotes Celtiform Stelae are themselves naturally derived from Olmec antecedents, the effigy celts, while the text of Stela C completely lacks such Olmec antecedents. With no Olmec antecedents, the only logical precursors to the crudely rendered and unfinished hieroglyphic text of Stela C are earlier Early Maya texts such as those which are found in the Guatemalan Piedmont! Finally, while the inscription of Tres Zapotes Stela C looks to the Pacific Piedmont for its antecedents, it (along with the Chiapa de Corzo ceramic fragment) provides an antecedent for the non-Olmec Tuxtla style inscriptions found later on the Tuxtla Statuette, the La Mojarra Stela, an unprovenanced serpentine mask (Winfield Capitaine 1988), the Cerro de las Mesas Stelae and the recently discovered El Alemain Stela.
The study of Celtiform Stela shows that, from at least the Middle-Preclassic, the Olmec shared in the Maya conception of celts and stelae as related objects. The evolution of the purely Olmec Celtiform Stela monument type suggests a lengthy popularity for the Olmec version of the celt as stela. The abscence of writing on Celtiform Stelae, or any other Olmec Monument type, indicates that writing was introduced to the Olmec from the Maya region. The abscence of writing in the "Olmec Heartland" prior to the arrival of strong Maya influence further indicates that, far from giving birth to Maya civilization, Olmec civilization was an illiterate contemporary which was ultimately overwhelmed and replaced by expanding Maya influence.
