Chapter 5

The 0lmec Dawning

               Momentous as the developments of the Ocós period had been within Soconusco, other regions of Pacific Mexico were also being brought increasingly into the orbit of the expanding civilizations of South America during this same time frame. In the alluvial lowlands of what today are the states of Colima, Jalisco, and Nayarit, the so-called Capacha ceramic complex had appeared about 1450 B.C. One of its hallmarks was a type of stirrup jar whose origins may be traced to the Machalilla culture, another variant of the Chorrera culture from Ecuador. From its toehold on the western coast of Mexico, the Capacha culture managed to penetrate into the interior of the country by following the valley of the Río Grande de Santiago, and sometime around 1300 B.C., distinctive chamber burials -- again suggestive of South American design -- were being cut into the volcanic ash as far inland as the upland site of El Opefio in Michoacán state.

               Returning once more to Soconusco, we find that archaeologists working in that region have set off the century and a half between 1350 and 1200 B.C. as a distinct phase in the cultural evolution of the area -- a period which they have called the Cherla. The distinguishing characteristic of this period they have likewise identified as "Olmec influence."

               The very mention of the Olmecs prompts us to question: Who were these people, and where did they come from? Certainly, from the time of the discovery of the very first "Colossal Head" in southern Veracruz state in 1862 until well into the 1980's, most of the finds which corresponded to an Olmec art style could be traced to what has been called by Ignacio Bernal "the Olmec metropolitan area," located in the Gulf coastal plain of Mexico. Naturally, this would imply that if the Olmecs originated in the general area of Veracruz and Tabasco states, they would have had to have moved southward across the Isthmus of Tehuantepec to have had any appreciable influence on the cultural evolution of the Soconusco region. However, such a postulated movement, especially at the time of the Cherla period, is totally unsupported by any archaeological evidence. On the other hand, there are numerous indicators that a vigorous movement in the opposite direction -- toward the north -- was going on at precisely this time, in which case the so-called Olmec influence must have been a native-born development emanating from Soconusco itself

               One such indicator is the Ocós ceramic style (Clark prefers to call it the "Locona"), which appears to have had a major northward diffusion at this time. Another indicator, less exactly datable to be sure but falling well within the time frame in question, is the degree of differentiation which has taken place between two branches of the Mayan language the tongue that was spoken throughout the Gulf coast region of Mexico as least as far back as 1500 B.C. Morris Swadesh, a renowned proponent of glottochronology, or the dating of languages by their relative differentiation from each other, has argued that Huastec, which is now spoken in the far north of Veracruz state, was separated from Yucatec, which is the indigenous language spoken in the area of Campeche and the Yucatán Peninsula, sometime during the thirteenth century B.C. Although he does not suggest a mechanism for accomplishing such a "split," the most likely possibility would have been the driving of a "wedge" into the Gulf coast region by a non-Mayan-speaking group. Such a wedge, in turn, would far more likely have been the result of a sustained overland movement from the south than an episodic seaborne invasion from the north. The continued presence of peoples of Zoquean speech at the northern end of the Tehuantepec Gap (the Popolucas in the eastern foothills of the Tuxtla Mountains), in the Oaxaca highlands to the west of the Gap (the Mixe), and in the Chiapas highlands to the east of the Gap (the Zoque) -- all of whom were linguistically related to the recently extinguished Tapachulteca dialect spoken in Soconusco -- testifies to the existence of a language whose geographic distribution once bridged the Isthmus of Tehuantepec from the Pacific to the Gulf.

Figure 13.

The diffusion of Ocós/Locona pottery out of Soconusco was no doubt paralleled by a diffusion of sculptural techniques as well. This contention is further strengthened by the realization that the absence of any stone whatsoever in the alluvial plains of eastern Mexico means that a tradition of stone carving can scarcely have arisen spontaneously there. When all such evidence is combined with what we know about the origins of the calendar, it becomes apparent that the conventional theories of 0lmec diffusion outward from the Gulf coastal plain must be turned quite around.


               In the field season of 1983, 1 and a student assistant journeyed into the backcountry of southern Mexico to visit and record the languages of the three extant groups of Zoque-speakers. Using a100-word diagnostic glossary devised by Swadesh, we attempted to make at least a crude statistical analysis of their degree of interrelationship. For whatever merit it may have, our analysis revealed that about two thirds of the words in the glossary were either shared by or very similar to one another in the Popoluca and Zoque languages, but that only about half of the words were recognizably similar between the Mixe and Zoque tongues. Of the three peoples, the Mixe were decidedly the most aloof and suspicious -- traits that likewise correlated very well with their high degree of isolation.

               However, at the time we made this journey into the backcountry of Mexico, I was unaware of the study made in 1976 by the linguists Campbell and Kaufman. They had concluded that the "Olmecs at least in part were Mixe-Zoque speakers" and that it made no difference to their hypothesis whether the Olmec origins were in the "Veracruz,, Tabasco, Morelos, or northeastern Oaxaca" areas (Campbell and Kaufman, 1976, 89-99). They were able to trace about 450 loan words in adjacent tongues which clearly had come from Mixe-Zoque and which demonstrated that the donor people had had a fairly sophisticated culture already around 1500 B.C. Not only were there many basic cultigens on that list, but also included were suggestive terms implying a vigesimal counting system, a calendar, measurement, divination, human sacrifice, "some form of writing," and the names of creatures such as the alligator and iguana. Perhaps constrained by the conventional view that the Olmecs had emanated from the Gulf coastal plain (or some region adjacent to it), Campbell and Kaufman expressed surprise that such "distant" languages as Xinca and Lenca in Central America also showed a strong Mixe-Zoque influence; however, for anyone recognizing Soconusco as their original core area (as I have argued here), finding their influence among the adjacent peoples in El Salvador and Honduras comes as no surprise at all.

               The evidence seems clear, therefore, that sometime in the Cherla period, an extensive movement of Zoque-speaking peoples spread out of Soconusco, across the Tehuantepec Gap and into the Gulf coastal plain of Mexico, bringing with them their characteristic Ocós-emulated pottery. Not only was this movement forceful enough to send the original Maya dwellers of the region scurrying for safety both to the north and to the east -- the Huastecs and the Yucatecs, respectively -- but its cultural characteristics were so distinctive that for the first time in the panoply of Mesoamerica's evolution we can speak of the appearance of an Olmec art style. Coming so closely in the wake of the development of both the sacred almanac and the secular calendar, we can ask: What prompted this explosive northward expansion? Was it the result of some great natural disaster in Soconusco, such as a volcanic eruption, a cataclysmic earthquake, a devastating hurricane, or a monstrous tsunami? Or some terrible internecine conflict which forced thousands of refugees to flee for their lives into the sparsely inhabited jungles of the north? Perhaps the explanation is that even resource-rich Soconusco had been pushed to the limits of its "carrying capacity," and overpopulation had finally obliged the Zoques to expand into the territories of their Maya neighbors to the north. Or, might the recently acquired calendars, with their implicit recognition of the importance of time and its fundamental role in religious life and ritual, have impelled them to launch a missionary "crusade" into these outlying regions? Certainly, any of these scenarios is possible, though some are more probable than others. All that we can be sure of is that the migration did in fact take place; what triggered it, we will probably never know.

Figure 14. To explain the pattern of distribution of native languages which existed in southern Mesoamerica at the time of the Spanish conquest, it is necessary to look back at the cultural geography of the region nearly three millennia earlier. At that time it appears that three distinct language families could be identified in the region: Along the entire Gulf coast, Mayan was spoken, whereas in the highlands of Oaxaca to the southwest, Zapotec was the principal tongue. In Soconusco the dominant language at the time appears to have been proto-Mixe-Zoque.

Figure 15.

According to Swadesh, about 3300 years ago the Mayan language family was split into two groups -- the Huastecs along the northern Gulf coast and the remainder of the Mayan-speaking peoples to the south and east. The author hypothesizes that this was the result of an overland invasion across the Tehuantepec Gap by Zoque-speakers, for this would have been at precisely the same time that the sacred calendar was being diffused from Izapa to the so-called 0lmec centers of the Gulf coast. The diffusion of language was just one of several cultural traits which appear to have taken place at the same time.

Figure 16.

The vigorous expansion of the Nahuatl-speaking Aztecs into southern Mesoamerica in the fifteenth century A.D. reshuffled the linguistic patterns of the region considerably. As the Aztecs pushed south to open trade routes to Soconusco, the Zoque-speaking peoples fled into the adjacent mountain areas to take refuge. As a result, people identified today as Popolucas inhabit the eastern foothills of the Tuxtlas, the Mixe withdrew into the highlands of Oaxaca, and the Zoque pulled back into the uplands of Chiapas. The original Zoquean core of Soconusco was all but submerged by Nahuatl-speakers, although the Tapachultecas managed to retain their own linguistic identity in the mountains back of Izapa until the latter decades of the nineteenth century.


               Although no archaeologist or linguist that I am aware of has ever hypothesized such a migration as that which I have described here, the geographic realities of its having taken place are virtually indisputable. No other single event or series of events can as logically explain (1) the separation of the Huastecs from the rest of the Maya family, (2) the scattered, refugelike distribution of Mixe-Zoque speakers in the mountains of southern Mexico, or (3) the presence of Mixe-Zoque influences among the peoples of El Salvador and Honduras. While I defer entirely to Swadesh as to its timing, I am happily reassured by how closely his glottochronology accords with my own dating of calendrical diffusion. By the same token, I am heartened that many respected archaeologists have come to believe that the origins of Olmec high culture may ultimately be traced to the Soconusco region (Adams, 1991, 43).


Formula 1: "The mountain where the sun turns"

               The earliest known ceremonial center which has been credited to the Olmecs is San Lorenzo, dating to ca. 1350 B.C. -- i.e., just at the break between the Ocós and Cherla phases of Soconuscan cultural development. San Lorenzo has intrigued archaeologists with its many sophisticated features: the artificial terrace some 50 m (160 ft) high on which it is built (which had been fashioned, some researchers claim, in the shape of a flying bird); its extensive stone-lined drainage system; and its numerous sculptures, many of them "Colossal Heads" which appear to have been toppled into adjacent ravines and ritually buried. To the geographer, though, its location is of special interest.

               Location, of course, may be defined in several different ways. For example, it can be expressed mathematically, in terms of latitude and longitude, but in the context of the Olmecs, who were unfamiliar with such concepts, such a definition would be meaningless. Location can also be expressed in relative terms, that is, one place may be defined in relationship to another. Using such a definition, the location of San Lorenzo, relative to the Zoque homeland in Soconusco, is seen to be adjacent; in other words, if the region of Soconusco itself was central to the life of the Zoques, then San Lorenzo was the nearest major Olmec ceremonial center to that area. Thus, its geographic proximity to Soconusco, together with its being the oldest known Olmec center, strongly suggests that it was founded as part of the northward expansion of Zoque-speaking peoples in the fourteenth century B.C.

               In speaking of urban settlements, of which San Lorenzo certainly was an incipient example, the geographer may employ two further definitions of location as well. One is site -- the description of the actual plot of ground on which the settlement was founded -- and the other is situation -- the relationship of a place to its surroundings. The former definition, for example, involves defining a place in terms of being located on a river terrace, on an island, at the top of a hill, or in the midst of a swamp. Although the terms used may be generic, such as terrace, island, hill, and swamp, the actual site of any given place is specific and unique. Hanover, New Hampshire, for example, is located on a postglacial river terrace known as the Hanover Plain, situated on the left bank of the Connecticut River, just above the mouth of Mink Brook -- a totally specific site, occupied by no other settlement. New York (or originally New Amsterdam) was located at the southern tip of Manhattan Island, where the Hudson River empties into New York Bay, whereas Venice, Italy, was founded in the marshes just to the north of the delta of the Po River on the coast of the Adriatic Sea.

               Defining a place's situation is to spell out its relationship to its surroundings. For example, Hanover, New Hampshire, has one state highway traversing it north-south along the Connecticut Valley and a bridge linking it with Vermont across the river; consequently, it does not command or interact with a very large hinterland, and as a result has far fewer commercial and industrial establishments than does a population node just five miles away that serves as the intersection of two railroads, two interstate highways, two federal highways, and two state highways, and that furthermore boasts a regional airport. In contrast, to describe the situation of New York City is to define the largest, busiest, most populous and extensive hinterland of any metropolitan area in the United States. In a comparative sense, Venice, despite its poor site (which has resulted in the city's sinking into the marsh), had a premium situation during the Middle Ages when most of Europe north of the Alps formed part of its hinterland, thanks to the easy access provided by the Brenner Pass.

               The comparison of Venice and New York makes another point, namely that the significance of places changes through time. When Venice was at the peak of its grandeur, there was no such place as New York. And there is nothing in the cards that says that New York will forever remain the most important urban agglomeration on the North American continent. By the same token, the decision of the Olmecs about 1350 B.C. to locate San Lorenzo precisely where they did was obviously colored by concerns which were important then but may be all but forgotten today.

               However, before we examine the specifics of San Lorenzo's location we should address a more general question relating to the absence of Olmec settlements between Izapa and San Lorenzo: Why were there no ceremonial centers founded in the intervening areas of Soconusco and Tehuantepec? Perhaps the most striking reason for this relative void in settlement was the local geography. As one moves northward through Mexico along the coast of Soconusco, not only does the coastal plain grow narrower but it also becomes considerably drier. By the time one reaches the vicinity of Tonalá, the foothills pinch close to the lagoons, the rain forest degrades into scrub, and the rivers have shriveled from perennial watercourses into intermittent streams. There simply was no hinterland capable of supporting an urban center of any size, at least in terms of the technological levels of the Olmecs. By the time the Pacific coastal plain opened into the southern approaches of the Tehuantepec Gap, the countryside had degenerated into a hostile semiarid region of low, thorny trees. As a people native to the rain forest, they did not again encounter an environment in which they felt "at home" until they had crossed over the divide of the Tehuantepec Gap into the vast tropical lowland of Mexico's Gulf coast. There they found a region of huge, meandering rivers winding their way through a canopy of lush green forest many times the size of their original home in Soconusco -- a beckoning land with whose ecological challenges and possibilities they were already thoroughly familiar.

Figure 17.

The climatic station of Villahermosa is typical of the Gulf coastal plain of eastern Mexico, the region Bernal has termed the "Olmec metropolitan area." Its water need, or temperature curve, varies a little more than that in Soconusco, because the low-sun months are somewhat cooler than on the Pacific coast. This is due to the outbursts of cold air ("northers") that sweep down from Texas at this season, bringing some rain. As spring arrives and the northers cease, a dry season is experienced until May when the summer monsoonal rains begin. The peak precipitation in the Olmec region occurs in the early autumn when the passage of a hurricane or two can bring especially heavy rains to the Gulf coast. Again, the warmth and moisture indices reveal that this is a tropical humid region whose native vegetation was a dense forest cover. (Data from Secretaría de Recursos Hidráulicos.)


               But how was it that they opted to build their first great ceremonial center at San Lorenzo? For one thing, the site which they chose was on the south bank of a branch of the Río Coatzacoalcos, the principal river draining the Gulf coast lowlands at the northern entrance to the Tehuantepec Gap. The lack of stone in the alluvial lowlands of eastern Mexico meant that the Olmecs had to quarry whatever building material they needed in the Tuxtla Mountains of southern Veracruz state where local volcanism had produced two or three stratovolcanoes and a profusion of cinder cones. (It is generally thought that the Olmec quarries were located near Punta Roca Partida on the north coast of the Tuxtlas.) From these seaside quarries, they would have had to freight the huge basaltic blocks from which they carved the "Colossal Heads" about 100 km (60 mi) along the coast before entering the river and poling them 50 km (30 mi) farther into the interior. Finally, they had to choose a location as close to the river bank as was practicable, for it was one thing to maneuver a 20- to 25-ton block of stone on a log raft but quite another to move it overland any distance.

               What, then, was so special about the site of San Lorenzo that they felt obliged to freight the stone 50 km (30 mi) upstream? Would not a similar terrace, closer to the ocean, have sufficed just as well? Moreover, even though the water levels in the great rivers of eastern Mexico may rise and fall by as much as 10 m (35 ft) between the wet and dry seasons, was it really necessary to construct a terrace 50 m (160 ft) high on which to build their ceremonial center? Clearly, if ease of transportation and protection from the river at flood stage were their primary concerns, a lower site closer to the Gulf of Mexico should have fulfilled their needs just as adequately.

               If the explanation for San Lorenzo's location cannot be found in its site, perhaps we should examine its situation instead. What is the relationship of this particular location to its surroundings? Just as Lucien Lefebvre pointed out that the primary motivations for early people to engage in travel were commercial, religious, and military in nature, so these same motives prompted them to found settlements. San Lorenzo's situation at the northern end of the Tehuantepec Gap certainly would have been a strategic one in terms of the exchange of goods -- had a commercial economy existed at the time -- but given what we know of the early Olmecs, that is highly unlikely. On the other hand, San Lorenzo could likewise have had a strategic importance in the military sense as well, for it did in effect control a major route of movement; its site, however, had surely not been chosen with any defensive considerations in mind, so such an explanation must also be dismissed. Essentially, we are left with a religious motive for its founding: But because its site had no special "sacredness" associated with it -- as did, for example, the oracle of Delphi, or the Kaaba in Mecca -- only its situation could have had a religious significance. What might this have been?

               Looking back into the Olmec experience -- as we understand it -- for some association between place and religious significance, the only example we have is that afforded by the siting of Izapa in relationship to the summer solstice sunrise over the volcano Tajumulco. That example is instructive in itself, because it suggests the central importance of the calendar, and, in this case, even of the secular calendar, in the lives of the people. Surely, as the Olmecs began their northward expansion out of Soconusco, it would have been readily apparent to the priests that, even though the sun continues to make two zenithal passages across the Gulf coastal plain in a given year, it does so at a shorter interval than in their homeland. Therefore, the 260-day count was no longer valid, nor were such key dates as "1 Alligator" or "1 Snake".  On the other hand, the sun continued to make two "turns" in the heavens -- once in the north and again in the south, at opposite ends of its annual migration. Although the zenithal passages no longer took place on the same days as they did in Soconusco, the days when the sun "turned around" did remain the same. Only now, because the expansion of the Olmecs had taken them across the Tehuantepec Gap and into the Gulf coastal plain, there were no longer any topographic features on the northern horizon against which to calibrate its extreme position in that direction. The mountains all lay behind them to the south, and it was quite apparent that the highest peak which they passed on their northward trek -- Zempoaltepec (3396 m, or 11,138 ft) -- still could be seen on the southwestern horizon. If, then, they were to retain a working version of the calendar which had been developed in Izapa, it would have to be based not on the zenithal passage of the sun but on its turning point in the heavens. Furthermore, because the sun's northern turning point could not be calibrated due to an absence of mountains in that direction, it would have to be its southern turning point instead. Moreover, because the highest mountain within view lay to the southwest, it would have to be calibrated at sunset rather than at sunrise as it was in Izapa. In short, whether they realized it or not, they would be seeking an alignment which was just the opposite of that at Izapa -- an orientation to a winter solstice sunset position rather than to a summer solstice sunrise.

               (A word about solsticial alignments in Mesoamerica: Even though the region extends through 10 degrees of latitude -- from about 13 to 23º N -- because it lies so near the equator, sun angles scarcely differ by more than a degree across the entire region, in other words, less than could be distinguished by someone practicing naked-eye astronomy as these people did. For all intents and purposes, the azimuth of the summer solstice sunrise can be equated to 65º throughout the region, or 25º north of east. Similarly, the winter solstice sunrise can be equated to 115º, or 25º south of east. Sunset positions on each of these days may be marked against the corresponding positions along the western horizon -- i.e., 295º, or 25º north of west for the summer solstice, and 245º, or 25º south of west for the winter solstice. Naturally, none of the Mesoamerican peoples reckoned in terms of angles or degrees, so although we express such measurements in these units, the alignments which they established were done solely through repeated observations in the field. It is, however, not inconceivable that, having once realized what the extreme points of the sun were, they made a graphic representation of its limits. If so, the resultant diagram would have approximated a recumbent cross -- a so-called Saint Andrew's cross -- which indeed has been recognized as one of the most frequently repeated artistic motifs used by the Olmecs.)

               Having already had the experience of locating the site of Izapa by using solsticial orientation as a "principle," the 0lmec priests were well prepared to do it again in the case of San Lorenzo. At Izapa the site that was chosen was the point where the alignment with Tajumulco (i.e., sunrise at the summer solstice, or an azimuth of 65º) intersected the Río Izapa, which served as a source of water for the ceremonial center. In the case of San Lorenzo, the site would be dictated by the point where the alignment with Zempoaltepec (sunset at the winter solstice, or an azimuth of 245º) intersected the Río Coatzacoalcos, not only for access to drinking water but also for expediting the transport of stone from the Tuxtlas. Naturally, pinpointing such a location may have taken more than one "field season," but even so, it was probably carried out by three or four "survey teams" rather than just one. The priest could initially have gotten a pretty good idea of the stretch of the river along which the proper vantage point for his planned ceremonial center would be found, and then, as the winter solstice neared, he would have posted his teams at intervals along the appropriate part of the waterway to observe the sunset over the distant mountain. The final choice of the site would be that of the team which reported the sunset directly behind the mountain.

Figure 18.

On this computer-generated map, the theoretical radius of visibility of certain key topographic features as viewed from sea level is shown. However, there is no guarantee that any given place within the radius of visibility has a clear line of sight to the topographic feature in question. This must be determined from either observations in the field or measurements made on large scale maps.


               That some version of the scenario just described must have taken place about the year 1200 B.C. seems quite certain, for large-scale maps reveal that the relationship between the site of San Lorenzo and the highest mountain within the range of its visibility, Zempoaltepec, is precisely that of the winter solstice sunset. Although there has been some debate over why the Olmecs had built their ceremonial center on an artificial terrace some 50 m (160 ft) above the river, it may well have been not only to protect their settlement from the high waters of the river in flood but also to improve their visibility of the mountain they had chosen as a calendrical marker. In any event, here we have exquisite evidence of a sophisticated idea having served as the locational. rationale for the earliest known ceremonial center of the 0lmecs -- a planned settlement whose location was a carefully chosen compromise between the need for drinking water, a navigable waterway, and the dictates of the sun-god!

               Here was urban planning on a grand scale. Whether the 0lmecs pressed local Maya into service in building the terrace on which San Lorenzo was to be located, or in quarrying and freighting the stone used for the sculptures of the "Colossal Heads," or in constructing the elaborate drainage system which served the ceremonial center is really beside the point. Here was a well-organized hierarchical command system at work, directed by priests who were thoroughly conversant with the calendar and its uses, who could allocate labor and resources to the most extensive manmade undertaking the region of Mesoamerica had yet seen, and all of it most probably in the name of religion. In a word, the Olmec civilization had been launched -- and in a most spectacular and dramatic manner.

               Yet, in the hindsight of history, something must have gone wrong at San Lorenzo, for by 900 B.C. it seems to have been abandoned. Perhaps internal dissension brought about a dissolution of its social structure. Perhaps what might have been a site pleasing to the sun-god was not especially conducive to meeting the daily needs of the common people; after all, the rainforests surrounding San Lorenzo lacked the richness and diversity of ecological resources found in Soconusco. Whatever it was, the great "Colossal Heads" which had been quarried, transported, and carved so laboriously were now rolled down into the ravines which dissected the edges of the terrace and carefully covered up, and within a generation or two, the jungle had totally reclaimed the site.

               This, of course, was not the undoing of the Olmecs, for already before 1000 B.C. a second ceremonial center, known to the modern world as La Venta, had come into being to the northeast. This time a site had been chosen much closer to the sea, in fact barely 20 km (12 mi) inland from the Gulf. Again, it lay on the banks of a navigable river, this time the Río Tonalá, which today forms the boundary between the states of Veracruz and Tabasco. And this time it had a site rather similar to that of Venice, in that it lay in the midst of coastal marshes. In fact, Michael Coe has picturesquely called it "a sanctuary in the swamps." However, this description is probably a more accurate characterization of Venice than it is of La Venta, for the former's founding was the result of people fleeing before the invading Germanic tribes, seeking shelter in the marshes of the Po delta, while we have no evidence of a similar episode having been responsible for the rise of the latter. On the other hand, the limited size of the island and its relative isolation in the swamps suggests that the principal function of La Venta may have been as a religious pilgrimage site (Adams, 1991, 62).

Figure 19. Most of the principal 0lmec ceremonial centers in the so-called metropolitan area of Mexico's Gulf coastal plain demonstrate solsticial orientations to the highest mountain within sight. For example, San Lorenzo is oriented to the winter solstice sunset over Zempoaltepec, Laguna de los Cerros is oriented to the summer solstice sunrise over Cerro Santa Martha, and Tres Zapotes has a similar orientation to Volcán San Martin. Interestingly, La Venta appears to be oriented both to the summer solstice sunset over Volcán San Martín and to the August 13 sunset over Cerro Santa Martha. As explained in the text, La Venta illustrates an evolution in locational principles that probably took place sometime about 800 B.C.


               Indeed, we have every reason to view with a certain degree of puzzlement just why the Olmecs chose this site for their second exercise in urban planning. Although it lies on the right bank of the Río Tonalá and is surrounded by marshes, its immediate setting is high and dry enough to have spared it from inundation by rainy-season floodwaters. (Whether its site has been artificially built up in any way has never been commented on in the archaeological literature; it is called simply an "island.") Certainly its commanding feature is a clay pyramid, or mound, some 30 m (100 ft) in height. Otherwise, its outstanding characteristic is its carefully planned symmetry around a central axis which is oriented 8º west of north.

               The latter feature has intrigued archaeologists but has never satisfactorily been explained by them. One attempt by Marian Hatch, arguing for the site's orientation to the star Eta Draconis (magnitude 2.73), sought to invoke an astronomical connection, but ultimately not a very convincing one (Hatch, 1973). In the heyday of La Venta's existence, about the year 1000 B.C., Eta Draconis would have set just shy of an azimuth of 351º, but thereafter, due to precession, or the "wobble" of Earth on its axis, it would have set farther and farther away from this point. (Indeed, in the same year its declination was 69º, or 21º away from the north celestial pole, which is the closest it has been within the last 3000 years.)

               On the other hand, anyone living in La Venta would have an absolutely unencumbered view of the entire northern horizon, having nothing but a featureless coastal plain merging with the sea in the distance. Surely, one of the first discoveries a skygazing priest would have made -- if he was not already aware of it -- was that the sky and/or its stars rotated. Here, at La Venta, he could see many of the constellations with which he had been familiar at Izapa, or perhaps even at San Lorenzo, make a complete circle in the heavens. Although he was already familiar with the fact that the same constellation had a different location in the sky at different times of the night, and for that matter, at different times of the year, once they disappeared over the western horizon and the sun had risen in the east, he had no idea where the stars had gone. Now he could see that they really went around in a circle.

               There is no evidence that this discovery in any way helped the early Mesoamericans to stumble onto the recognition of Earth's sphericity or of its diurnal or annual movements. But what it did do was to cause the skygazing priest to zero in on that part of the sky which moved the least. In the year 1000 B.C., the star which came closest to being at the "center" of the sky was Kochab (magnitude 2.07), whose declination was 83º.5. Since Kochab is circumpolar at the latitude of La Venta, there would be no way that the skygazer could fix its position against the horizon; therefore, if he had oriented the site to what he presumed to be its farthest west position, he may well have misjudged the alignment of La Venta's central axis by about 1º.5. This is still an impressive approximation, given that he most likely had no instruments other than a pair of crossed sticks to aid him in making his observations.

               As it turned out, the Olmec propensity for order and symmetry definitely worked to the archaeologists' advantage in excavating the site. After Matthew Stirling had discovered La Venta's axial arrangement aimed at an azimuth of 352º and uncovered a huge mosaic depicting a jaguar's face made out of carefully quarried blocks of green serpentine, Robert Heizer and Philip Drucker, who continued the site's excavation in the 1950's, correctly surmised that another such mosaic must lie buried on the opposite side of the axis. The twin mosaics, which had been ritually covered by layers of different-colored sands, were composed of some 1200 tons of stone which had been laboriously cut and hauled from Niltepec, some 180 km (110 mi) to the south across the Tehuantepec Gap. La Venta, often called "the capital of the Olmecs," was not only another triumph of the Mesoamerican intellect, but also a vivid illustration of the organizational skills of a young and dynamic civilization. It appears to have functioned as the most important center of the Olmecs at least until about 600 B.C.

               Although the internal layout of the site itself can possibly be attributed to an orientation toward Kochab, the specific siting of La Venta has yet to be explained. Why was it located precisely where it was on the banks of the Río Tonalá? Certainly, the access to higher and drier land, drinking water, and a navigable river were imperative, because La Venta had to import its building stone from the Tuxtlas in the same way that San Lorenzo did. Was the 30-m- (100-ft-) high pyramid simply another manifestation of the triumph of political organization, or did it have a utilitarian function as well? Because La Venta clearly lacked any strategic advantages, either in a commercial or a military sense, the choice of its site must have been dictated even more strongly by a religio-astronomic motive.

               If one climbs the pyramid on a relatively clear day, mountains are visible on the horizon to the southeast in Chiapas, to the southwest in Oaxaca, and to the northwest in Veracruz state. However, it is by far the latter -- the Tuxtlas -- which are the closest, and hence the most prominent of such features.

               Three stratovolcanoes dominate the skyline of the Tuxtlas. The nearest to La Venta as the crow flies is Cerro San Martín at a distance of 76 km (47 mi). Almost directly behind it, as seen from La Venta, stands Cerro Santa Martha at a distance of some 88 km (55 mi), and back of that in turn rises Volcán San Martín at a distance of 132 km (82 mi). Although various map sources give differing elevations for these three peaks, the nearest one (Cerro San Martín) is the lowest (1250 m, or 4100 ft) but nevertheless the most visually prominent from La Venta.  Judging from its configuration, it appears to have been considerably higher in the past, for the long gradual slopes of what once was its cone terminate abruptly on their inner sides, suggesting that the entire top of the mountain has been blown away. Whether a catastrophic eruption of such massive proportions occurred within the 3000 years since La Venta was founded, it is impossible to say at this time. (Evidence of a volcanic eruption of absolutely cataclysmic proportions has been demonstrated for Ilopango in El Salvador about the year A.D. 260, so such a possibility is not as "fanciful" as it might first appear.) On the other hand, Cerro Santa Martha is without question the highest of the existing peaks in the Tuxtlas. Depending on which map one consults, its elevation is given as 1787 m (5861 ft) or as 1879 m (6163 ft). It is clear from the most detailed Mexican map of the region (compiled from aerial photographs taken in 1976) that Cerro Santa Martha's crater has been extensively breached on its northeast side by vigorous stream erosion, but again, how much of this has occurred within the last three millennia is not easily determined. In any case, local informants note that Cerro Santa Martha has water in its crater (suggesting that it may have been dormant for some time) and contrast this with Cerro San Martín's crater, which they claim has sulfur in it. In any event, the azimuth measured between La Venta and the highest remaining lip of each of these craters is slightly over 285º, or a good 10º away from the sunset position on the summer solstice. On the other hand, the most distant of the three stratovolcanoes -- Volcán San Martín, whose height is variously given as 1402 m (4600 ft) or 1773 m (5815 ft) -- lies at an azimuth of ca. 295º, and therefore may well have served La Venta as a topographic marker for the summer solstice sunset (see Figure 19).

               A third major Olmec site has only more recently been discovered in the Gulf coastal lowlands of eastern Mexico, but as yet has not been systematically excavated. Known as Laguna de los Cerros, it is situated on the southern edge of the Tuxtlas in the valley of the Río San Juan, a tributary of the northwesterly flowing Río Papaloapan. It is thought to be almost as old as San Lorenzo and possibly somewhat older than La Venta, an assumption which accords perfectly with a postulated Olmec diffusion spreading outward from the Tehuantepec Gap. Furthermore, its location is again a compromise between a water-oriented site with access to a river, both for the water itself and for the transportation it affords, and a situation which is aligned to the summer solstice sunrise over Cerro Santa Martha (see Figure 19).

 Figure 20.

By about 1000 B.C., knowledge of the calendars and the principle of solsticial orientation had spread into the Olmec metropolitan area. San Lorenzo at the northern entrance to the Tehuantepec Gap and La Venta and Laguna de los Cerros on the eastern and western approaches of the Tuxtla Mountains, respectively, were thriving centers at the time. The Olmec fascination for jade, which in Mexico is only found in the Balsas Depression of Guerrero state, may have prompted the founding of Teopantecuanitlán by that date as well. Discovered only in 1985, the site is currently being excavated by scientists from Mexico's Institute of Archaeology and History.


               The diffusion of the Olmecs and their calendrically-based religion was not solely northwestward into the heartland of present-day Mexico, however. There is also evidence of a very early spread of their cultural influences south and eastward into modern Guatemala, El Salvador, and the borderlands of Honduras and Nicaragua. Farther eastward in Soconusco lies Abaj Takalik, where artifacts of Maya origin are found superimposed on Olmec foundations and at least two early Long Count inscriptions have also been uncovered. Not too surprisingly, the site likewise seems to have been chosen for its alignment to the volcano Santa Maria at the summer solstice sunrise. Here a navigable waterway was not a concern, however, because basaltic boulders for construction abound in the immediate vicinity.

               About 80 km (50 mi) farther east yet is El Baúl with its rich assemblage of primitive sculpture (some of it magnetic), another very ancient Long Count inscription, and a strongly suggestive orientation to Volcán Agua at the summer solstice sunrise. We have already described nearby La Democracia as the center of the "Fat Boy" sculptures, with their intriguing magnetic properties.

               Certainly the largest of the Guatemalan sites with Olmec-like carvings is Kaminaljuyú, on the western outskirts of the country's capital city. Once a sprawling site of more than 200 platforms and pyramids, it has been severely damaged by Guatemala City's urban growth and few of its original structures remain intact. Dated to at least 800 B.C., and possibly earlier, Kaminaljuyú occupied a similar geographic situation to that of San Lorenzo in the sense that it commanded a strategic pass across the Atlantic-Pacific divide. Ironically, the belated Spanish recognition of this place's "centrality" within Guatemala caused them to relocate the colonial capital to this location from Antigua Guatemala, following the latter's damage in a severe earthquake in 1773. Like other "Olmecoid" ceremonial centers, Kaminaljuyú also demonstrates a solsticial orientation --in this instance, to the Volcán Fuego at the winter solstice sunset (see Figure 21).

               Perhaps the southeasternmost of the Olmec-influenced ceremonial centers of Mesoamerica was Tazumal, situated in what is the northwestern corner of present-day El Salvador. Not only are Olmec influences clearly seen in its artwork, but the site appears to have been oriented to the winter solstice sunrise over what formerly was the great volcano of Ilopango, just to the east of San Salvador, the country's capital. The latter volcano -- now a vast water-filled caldera -- experienced a massive eruption about A.D. 260, which devastated the adjacent settled areas and gave rise to large-scale out-migrations of the local people (see Figure 21).

Figure 21.

Olmec diffusion southward and eastward into the Pacific coastal plain of Guatemalan Soconusco and what is today the adjacent part of El Salvador saw the founding of several other early ceremonial centers, many of which continued the tradition of solsticial orientation to the highest mountain within sight.


Formula 2: "The fifty-second sunset"

               Sometime during its six to eight centuries of existence, La Venta witnessed the construction of a new series of structures (the so-called Stirling Complex) near the southern end of its great plaza. Interestingly, when these structures were first mapped, they were shown to have deviated some 23º.5 from the site's axis, which, as we have explained, is aimed 8º west of north. The fact that this angle corresponds to the inclination of the earth's axis is probably strictly a coincidence, but in any case, it means that the walls of the Stirling Complex are oriented 15º.5 off of the cardinal points. Thus, in a northerly direction it has an azimuth of 15º.5, in an easterly direction it faces 105º.5, to the south the azimuth reads 195º.5, and to the west, 285º.5. Only the latter of these azimuths is of any interest, because it marks the sunset position on August 13 -- "the day that time began," according to the sacred almanac. Could this alignment be an architectural "reinforcement" of a topographic orientation toward Cerro San Martín and/or Cerro Santa Martha? If so, it could mean that by about 1000 B.C., priests at La Venta had come up with a formula for recording when the zenithal sun was passing overhead at Izapa!

               (It should be noted that in a more recent survey of La Venta carried out by the National Institute for Archaeology and History [a copy of which is reproduced in Adams, 1991, 56 - 57], the structures of the Stirling Complex are shown to have the same axial alignment as the remainder of the site -- i.e., 8º west of north. If the original survey was in error, naturally the arguments presented above are no longer valid. However, this does not invalidate the discussion which follows for how the "formula" itself was derived.)

               In reality, the formula was as simple as it was ingenious. The problem at San Lorenzo had been that the priests had no way of knowing when it was August 13, because in their part of the world the zenithal passage of the sun did not occur on that date. Thus, they had settled on using one of the solstices instead, because the date of the sun's turning point was the same everywhere, they had discovered. Whereas at San Lorenzo they were obliged to use the winter solstice sunset to calibrate their calendar, when La Venta was founded it appears that they could once more think in terms of the summer solstice, as had originally been done in Izapa. Indeed, the only difference was that instead of marking the sunrise as they did at Izapa, they were obliged to use the sunset at La Venta.

               Once back in the mental groove of using the summer solstice to calibrate the secular calendar, it would not have been long before some priest realized that the beginning date of the sacred almanac can itself be calibrated by reference to the summer solstice. In effect, he was recognizing that, if the solstice occurred on June 22 and the "beginning of time" occurred on August 13, there was a fixed interval of time between these two dates. Using our modern calendar to demonstrate his thought process, we would count 8 days to complete the month of June, add 31 more for the month of July, and then count 13 until the sunset of August 13, yielding a total of 52 days. (For anyone used to thinking in "bundles" of 20's and 13's, what a neat package this was -- 4 rounds of 13 days = 52 days.) Thus, no matter where one wanted to build a ceremonial center, one could always find out when it was August 13. All that was required was to count 52 days from the time that the sun turns around in the north and mark the horizon at sunset!

               In pinning down the time frame of the discovery of this second principle of orientation, we can say that it was probably not in use when La Venta was founded about 1000 B.C. -- but only if Volcán San Martín had been the topographic feature against which the summer solstice sunset had been calibrated. If, however, the suggestive azimuth of either Cerro San Martín or Cerro Santa Martha had been employed in locating the site, then it would already have had to have been known at that date. In any event, it seems almost certain that it had been recognized by the time the site was abandoned about 600 B.C. If it were possible to date the founding of some other ceremonial site(s) that likewise had an August 13th alignment to within La Venta's span of existence -- i.e., between roughly 1000 and 600 B.C. -- then perhaps we could narrow down the timing of the discovery of this second principle more closely.

               Tres Zapotes, another major Olmec center located farther west along the coastal plain of Veracruz, is generally believed to have been founded about 800 B.C. Its orientation is to Volcán San Martin in the Tuxtlas, but it uses the mountain to mark the summer solstice sunrise (see Figure 19). A host of other ceremonial centers extending northward along the coast and westward onto the Mexican plateau likewise date to the general period between 1000 and 800 B.C. and invariably demonstrate solsticial alignments: The Totonac site of Remojadas, for example, is oriented to the summer solstice sunset over Nauhcampatépetl, or Cofre de Perote, as it came to be known to the Spanish. Similarly, Zempoala, farther north along the coast of Veracruz, is aligned to Citlaltépetl, or Orizaba, at the winter solstice sunset (see Figure 22). In the highland basin of Puebla, the pyramid of Tepenapa. at Cholula -- the New World's largest pre-Columbian structure -- is oriented to Ixtaccíhuatl at the summer solstice sunset. In the Valley of Mexico, the two oldest known agricultural settlements -- Zacatenco and El Arbolillo -- both of local origin, clearly had no such alignments, whereas Tlatilco and Tlapacoya, the most ancient of the Olmec-inspired settlements in the basin, are clearly oriented to the winter solstice sunrise over Ixtaccíhuatl. And near the southern edge of the valley, the four-tiered round pyramid of Cuicuilco, probably the oldest true ceremonial center in the basin of Mexico, is unmistakably aligned to the cone of Popocatépetl at the winter solstice sunrise.

Figure 22.

Farther north along the Gulf coast, we find that the Olmec site of Cerro de las Mesas (founded ca. 800 B.C.) is oriented to Citlaltépetl (Orizaba), the highest mountain in Mexico, on the August 13th sunset. Lacking a solsticial orientation, it may therefore be the first ceremonial center in Mesoamerica where the August 13th orientation was used. Both Remojadas and Zempoala, on the other hand, demonstrate solsticial alignments: the former to the summer solstice sunset over Nauhcampatépetl (Cofre de Perote), and the latter to the winter solstice sunset over Citlaltépetl.

Figure 23.

On the Mexican plateau, all ceremonial centers where "Olmec" origins or influence are found demonstrate solsticial alignments. The twin volcanoes of Popocatépetl and Ixtaccíhuatl were used as winter solstice sunrise azimuths for Cuicuilco, Tlatilco, and Tlapacoya in the Valley of Mexico and as a summer solstice sunset azimuth by Cholula in the Puebla basin. Ixtaccíhuatl also served to mark the winter solstice sunset from Tizatlán. One of the most intriguing solsticial alignments is that of Teotihuacán, which lines up perfectly with the winter solstice sunrise over Citlaltépetl (Orizaba). Because a low ridge intervenes between them, the mountain is not visible from the city; however, in 1993 what may have served as a "relay station" between the two was investigated by the author. Later Toltec ceremonial centers such as Tulancingo and Tula pointedly demonstrate that the principle of solsticial orientation was either forgotten or had died out by the time they were founded.


               However, the most innovative orientation along this northward prong of 0lmec advance seems to have been that of Cerro de las Mesas (most likely founded ca. 800 B.C.), about 60 km (35 mi) southeast of Veracruz. Unless we are dealing with merely another suggestive "coincidence," this site aligns to Orizaba at sunset on August 13. Therefore, if the date of its founding is accurate, we probably have here the first illustration of the application of the "summer solstice + 52 days" formula anywhere in the Olmec world -- at least outside of La Venta (see Figure 22).

               It should also be kept in mind that northward and westward beyond Cerro de las Mesas, the 0lmec influence seems to have been limited to the activity of a small elite composed of "missionaries" or "soldiers" or both, rather than any massive in-migration as seems to have been the case in the southern Gulf coastal plain. The fact that this elite could have had such a strong impact on the peoples of these other regions would seem to testify to the strength of both their religious and their intellectual message.

               Another prong of Olmec expansion -- no doubt of the same elitist militaristic-missionary type -- can be traced up the valley of the Río Tehuantepec into the highlands of Oaxaca where the principal "converts" were the Zapotecs. Their main ceremonial center, Monte Albán, dates back to at least 600 B.C., but differs from virtually all other such settlements in Mesoamerica in being located at the top of a mountain -- a 500-m (1600-ft) eminence which overlooks the present-day city of Oaxaca. With a commanding view over the great Y-shaped valley formed by the Río Atoyac and its main tributary, Monte Albán has a spectacular site, but one which could not have been dictated by any solsticial alignment. It would have been sheer chance if the mountain on which it is located happened to line up with another mountain -- and specifically the highest within view! -- on any of the key dates of the year. (Indeed, as the Olmecs penetrated the mountain fastnesses of the Sierra Madre del Sur, they appear to have had to abandon their predilection for orienting sites to commanding topographic features, because in most instances the very ruggedness of the region precluded the establishment of long lines of sight.) This means, therefore, that if any calendrical orientation is to be sought at Monte Albán, it would most likely commemorate August 13, as the beginning of the sacred almanac, and hence would have

Figure 24.

Among the earliest recognizable calendrical glyphs are these discovered at Monte Albán and thought to date to the time of the ceremonial center's founding (ca. 600 B.C.). However, evidence uncovered by archaeologist David Peterson strongly suggests that the calendar arrived in Monte Albán in fully developed form, because the date August 13 played a fundamental role in its annual cycle even though it had no local astronomical significance whatsoever.


employed the "summer solstice + 52 days" formula. If that had actually been the case in the initial layout of the ceremonial center, we will never know, because the great plaza at the top of the mountain underwent a drastic process of "urban renewal" about the third century B.C. when all the buildings save one -- the so-called Mound J, an arrow head-shaped structure near the south end of the plaza -- had been reconstructed on an axis 5º east of north. On the other hand, on a slightly lower ridge of the mountain to the northeast, a structure known as Mound Y (also identified as Tomb 105) clearly reveals from the large, irregular stones used in its construction that it had been one of the earliest buildings on the site, but due to its being located relatively off to the side, it had escaped the later "urban renewal" process. Significantly, when its orientation was checked, it was found to look out at the western horizon at an azimuth of 285º.5 -- the alignment of the setting sun on August 13. This discovery means that the Zapotecs were not only aware of the sacred almanac by at least 600 B.C., but that they had oriented at least one of their key structures to its beginning date, which indicates that the "summer solstice + 52 days" principle was already known at that time.

               Because of the antiquity of the inscriptions at Monte Albán, Alfonso Caso, among others, believed that the Mesoamerican calendrical systems had their origins in the highlands of Oaxaca. However, the recent work of anthropologist David Peterson at Monte Albán has revealed that "the 260-day calendar definitely did not originate at Monte Albán," but "must have been developed elsewhere and been brought to Monte Albán in a fully developed state" (unpublished manuscript; personal communication). Interestingly, his evidence suggests that the original Zapotec version of the 260-day calendar had days with fixed numbers and names, beginning on August 14 and running through April 30. Thereafter, the first 105 day-numbers and -names are repeated in a second, shorter cycle which ran from May 1 to August 12. Peterson maintains that August 13 was a nameless, or "zero," date which was probably never recorded, and that every four years it was repeated and/or counted twice, allowing the Zapotec calendar to remain synchronized with the true length of the solar year. Peterson also notes that the side walls of one of the oldest structures at Monte Albán, Mound K, are oriented to the sunrise positions on March 9 and October 5, which he suggests reflects the Zapotecs' method of defining a 52-day interval before and after the zenithal sun passages at Izapa. Both of these ideas -- a 365-day calendar with fixed day-names and divided into two separate cycles, one a longer 260-day cycle beginning on August 14 and the second a shorter 105-day cycle beginning on May 1; as well as a means of architecturally establishing a 52-day interval before and after these two critical dates -- have, of course, special relevance to the discussion at hand.

               Thus, if the "summer solstice + 52 days" formula had not already been discovered by the time of La Venta's founding, it certainly seems to have been in use when Cerro de las Mesas was founded a couple of hundred years later, and from there it appears to have diffused both back to La Venta, where it may have been incorporated into the alignment of the Stirling Complex (if one accepts the site plan produced in the initial survey), and also to Monte Albán, where it had dictated the layout of Mound Y around 600 B.C. However, the special twists given to the sacred calendar and the means of calibrating its beginning and ending dates are Zapotec innovations of which we have only recently become aware, thanks to Peterson's research.

               Two other calendrical innovations also seem to be the product of the Zapotecs because there is no evidence that they occur anywhere outside their culture area. Peterson makes the point that Mound J, the arrowhead-shaped structure near the southern end of the main plaza, appears to have been constructed to commemorate the azimuth of the sunrise on the days of the zenithal sun passage over Monte Albán (May 8 and August 5), the alignment of its original front steps having been 72º. However, the "Point" of the arrowhead is aimed at an azimuth of 252º -- orientation which he claims commemorates the sunset azimuth on the days when the sun is at its anti-zenithal position, or in other words, when it is overhead at the corresponding latitude in the Southern Hemisphere (November 10 and February 2). It is interesting to note that the only other Mesoamerican structure having the same kind of arrowhead configuration and demonstrating a similar (supposed) anti-zenithal orientation is the so-called Mound 0 at Caballito Blanco, about 30 km (20 mi) to the east of Oaxaca. Sometime around the year 275 B.C., when Monte Albán underwent a major urban renewal, a massive platform was added to the front of Mound J, with its new staircase being oriented to an azimuth of 48º. Anthony Aveni has demonstrated that in the years surrounding that date, the star Capella (magnitude 0.97) rose at this azimuth just before dawn on the day of the first zenithal sun passage, i.e., May 8; thus, from having been a structure whose orientation was totally dictated by solar azimuths, Mound J from 275 B.C. onward incorporated a stellar alignment into its newly reconstructed front façade.

               For a people who were unfamiliar with the sphericity of the earth, or for that matter, with a heliocentric solar system, the notion of the anti-zenith seems to have been a quantum leap into the unknown. What possible significance such dates as November 10 and February 2 could have had for the residents of either Monte Albán or Caballito Blanco is difficult to imagine. To be sure, they bracket the winter solstice in the northern hemisphere by 89 days, just as May 8 and August 5 bracket the summer solstice by a similar interval, but aside from such temporal symmetry there is little else in common between these pairs of dates.

               I would suggest, rather, that it was not the southwestern point, or "arrowhead," of Mound J that was important, but the alignment of its southern and western walls, which form a right angle oriented to the cardinal points. Thus, although all the other structures surrounding the great plaza of Monte Albán have walls oriented about 5º east of north, this is the only part of any structure at the site whose architecture preserves a true north-south, east-west alignment. When viewed from either end of the plaza, the west wall of Mound J's arrowhead could have served as a meridian, remaining in shadow from dawn until local noon and then basking in full sunlight until the sun set. That such a use might have been made of this feature is attested by a series of photographs taken by the author in January 1993 revealing how the vertical wall literally "flashed" into light at local noon, i.e., 12:27 P.M. The western wall could also have served as a meridian marker by anyone viewing the structure from the opposite end of the plaza, and thus may have been used to calibrate midnight as well.

               Beyond the mountains to the northwest of Monte Albin lived a people who posed a constant threat to the Zapotecs and who ultimately moved in to conquer them. These were the Mixtecs, or so-called "people of the clouds," who, because of their geographic location, were not reached by Olmec missionaries and the innovation of the calendar until sometime after the Zapotecs. Indeed, archaeologists date their ceremonial center at Huamelulpan to ca. 300 B.C., and although not much remains of the main pyramid there, its massive foundation stones not only bear carvings of calendrical glyphs in virtually mint condition but they also are squarely oriented to the sunset position on August 13. Thus, both the calendar and the knowledge of how to calibrate it had reached into the mountain fastnesses of the Mixtecs by at least the third century before Christ.

Figure 25.

By 500 B.C., Olmcc influence had diffused the urbanization process both eastward along the Pacific Piedmont of Guatemala and El Salvador and northwestward into the highlands of Oaxaca and onto the Mexican plateau. Although places like San Lorenzo had already been abandoned, an urban network of over a dozen ceremonial centers stretched for 1000 km (600 mi) through the heart of Mesoamerica.


               Yet another prong of Olmec advance led south and westward out of the Puebla basin along the valley of the Río Atoyac into the Balsas Depression. The principal ceremonial center located near the Atoyac's headwaters was Las Bocas, not far from the present-day city of Izúcar de Matamoros. Its site appears to have been chosen to permit the calendar's calibration by observing the summer solstice sunrise over Citlaltépetl (Orizaba), located 135 km (84 mi) to the northeast. To the west of Izúcar, near the banks of the Río Amatzinac, lies the ceremonial center of Chalchatzingo whose site seems to have been chosen, in part at least, because of the spectacular volcanic plug on which much of its "monumental" art has been etched. However, there is the intriguing possibility that a calendrical concern may likewise have been involved in the choice of its location, for it is directly in line with the setting sun at the summer solstice over the volcano Nevado de Toluca (4624 m, 14,954 ft). The 0lmec site at Gualupita, now swallowed up by the modern city of Cuernavaca, likewise seems to owe its location to its orientation to the summer solstice sunrise over Ixtaccíhuatl, 105 km (65 mi) to the northeast. But probably the principal ceremonial center along this southwestward prong of 0lmec expansion was Teopantecuanitlán, or "the place of the temple of the jaguar," discovered in 1985 by Guadalupe Martínez Donjuán.

Figure 26.

The main pyramid at the Mixtec site of Huamelulpan, Oaxaca, was constructed about 300 B.C. The massive stone blocks which underlie its southwestern corner not only demonstrate some well-preserved calendrical glyphs but also reveal an orientation of the front wall to the sunset position on August 13.


               Strategically located in a broad valley at the confluence of the Balsas River and its largest right-bank tributary, the Amacuzac (also called the Río Grande), Teopantecuanitlán is built around a sunken court, two walls of which are surmounted by great jaguar-face monoliths. Although the site may have been occupied from about 1400 B.C. to 600 B.C., it appears that its principal structures date from ca. 1000 B.C. Here in the hot and arid Balsas Depression of northern Guerrero state, not far from where Covarrubias had argued that the 0lmec hearth was to be found, the combination of the rivers with their life-giving water and the fertile alluvial soils made for an agricultural oasis of great productivity. It is also likely that one of the principal sources of Olmec jade lay somewhere nearby in the Balsas Valley of Guerrero (Adams, 1991, 73), hence the reason for their early penetration of this remote region. On the other hand, both the reported antiquity of the site (supposedly predating the origin of the sacred almanac) as well as the ruggedness of the local relief in the adjacent Sierra Madre del Sur region (precluding any long distance lines of sight) essentially rule out a calendrical motivation for the choice of Teopantecuanitlán's location.

               Perhaps the most triumphal breakthrough of the calendar and the religious ideology that embodied it came in the Valley of Mexico, due primarily to a natural catastrophe which occurred there about 200 B.C. The volcano Xitle, located a short distance to the southwest of the ceremonial center of Cuicuilco, began an ominous eruption, sending streams of lava cascading down toward the ancient settlement. The priests and the people had plenty of warning, so there was nothing to do but evacuate, and this they did by rounding the shores of the great Lake of the Moon to the northeast, thereby putting that body of water between themselves and the fiery volcano. From the relative safety of their refuge on the northeast side of the lake they could watch as the glowing rock flowed down over their city, covering everything but the main pyramid -- a circular, four-tiered structure surmounted by a broad ramp. When the eruption was over and the lava had finally cooled, the only evidence that a vibrant, thriving ceremonial center had ever occupied the site was the deserted pyramid now locked in a ring of solid stone. (Were it not for this pyramid projecting up through the lava field, it would have been doubtful that anyone would ever have known that a city lay buried there. Unlike most other sites that are excavated a spadeful of earth at a time, Cuicuilco had to be dynamited to reveal the secrets of its past) (see Figure 23).

               Already the heirs of several centuries of history, the survivors of Cuicuilco were not about to surrender to fate just because they had lost their city. Almost immediately they set about laying out an even grander metropolis, conceived and constructed on an immense scale in the relative safety of a broad plain bordering on the Lake of the Moon and commanding a low pass out of the Valley of Mexico toward the lowlands of the Gulf coast. Not only did it occupy a strategic location for trade, but it also controlled an immensely valuable resource which it could exchange for products from other regions --namely obsidian, or volcanic glass, highly prized for making tools and weapons because of the razor like edges it yielded.

               We know the city only by the name the Aztecs called it: Teotihuacán, or "Place of the Gods." They gave it this name because of the colossal size of its pyramids, despite the ruined condition in which they found them many centuries later. Even today the Pyramid of the Sun and the Pyramid of the Moon loom out of the plain like miniature mountains. Thinking these pyramids to be the resting places of gigantic deities, the Aztecs called the broad avenue running down the middle of the site the "Street of the Dead." What the Aztecs did not know, and what remained for the archaeologists to discover when the site was photographed and surveyed from the air in the 1960's, was that the "Street of the Dead" was merely the principal axis of an urban settlement that sprawled over hundreds of acres. Meticulously oriented to a grid that was offset from the cardinal points by 15º.5, the city had impressed its pattern on the entire countryside, even, it appears, influencing the layout of the chinampas, or "floating gardens", at the far southern end of the Lake of the Moon. Of course, once this pattern was mapped, an explanation had to be sought for its orientation, and in the course of time several hypotheses have been advanced. The explanation that is given visitors to the site in the little local museum is that it defines the sun's setting position on the day that the zenithal sun passes over. For Teotihuacán, located at latitude 19º.5 N, that occurs on both May 18 and July 26. Using a fairly simple trigonometric equation in which we insert 19º.5 in the formula for both the declination (latitude) of the sun and the latitude of the place in question, we obtain a sunset azimuth of 290º.7 -- a full 5º.2 off of the alignment which the excavators of Teotihuacán themselves have measured.

               Several other attempts had been made to explain the layout of Teotihuacán by means of astronomy, including that of archaeologist James Dow, in 1967, who thought it had to do with either the setting of the Pleiades or the rising of Sirius. In 1973, astronomer Anthony Aveni settled on the hypothesis that the site had been oriented as it had because it marked the setting position of the Pleiades about the year A.D. 150, the date when he assumed the city had been laid out. However, he concedes that if his dating is off by even so much as 50 years, the alignment would be 0º.5 in error due to the effects of precession. (Of course, the same kinds of difficulties would be involved in any argument based on star positions.) To account for the host of other ceremonial centers which share what he has generalized as the "17º offset" of Teotihuacán, Aveni further assumes they were simply copied from the original, because by the time they were built -- many of them in the period after A.D. 1100 -- the Pleiades idea would be off by as large an error as the zenithal sun-passage idea had yielded. Ironically, he essentially demolishes his own thesis a little farther along in his paper by admitting that the Pleiades are so faint as they near the horizon that they can hardly be seen when they are setting anyway (Aveni, 1973). (Despite this inconsistency, when National Geographic published a feature article on Teotihuacán in its December 1995 issue, they once again cited Aveni's hypothesis as the explanation for the city's grid pattern.)

               A couple of years after Aveni's explanation had been advanced, two new solutions to Teotihuacán's orientation were put forward. One was the result of excavations at the site which revealed that beneath the Pyramid of the Sun lay a lava cave which opened to the northwest. Conclusion -- It had been the aperture of this cave which provided the alignment upon which the entire city was plotted with such care (Heyden, 1975). However, no archaeologist has ever explained how such a finely measured orientation could ever be derived from the jagged walls of a basaltic tube -- and they probably never will, because it can't really be done. Ironically, Millon and his disciples use the astronomical orientation of the cave as their raison d'etre for the city's layout, while at the same time appropriating the notion that the cave defined "the place where time began" (Berrin and Pasztory, 1993, 17-19). Of course, in the process they overlook the fact that the Mesoamerican "celebration" of both the August 13th sunset and the solstices had been going on for well over a millennium before Teotihuacán was even founded.

               It remained for me and a group of my students to rediscover the principle which had motivated the city planners of Teotihuacán some 20 centuries earlier. Sitting atop the Pyramid of the Sun one January morning in 1975, I asked my students to look out over the site and consider the layout of the city below us. If the structure on which we were sitting was correctly named -- and most prehistoric places or structures are not, as for example, Teotihuacán itself -- what was it oriented toward, I asked. Because we had just climbed up the steps which surmount the pyramid on the side facing the "Street of the Dead" (i.e., the west-northwest), they immediately answered, "To the sunset." I congratulated them on their perspicacity and then asked them, "On which day?" Of course, most of them had seen the explanation provided in the local museum, so the answer came back, "The days that the sun passes directly overhead." This time I chided them for having "cheated" by recalling what was said in the museum, and then I asked them to confirm this figure by making their own calculations. A few minutes later the group of disgruntled students turned to me and reported that "it doesn't work; those are not the days." "All right then," I continued, "on what days does the sun set directly opposite the Pyramid of the Sun?"

               Out came the calculators again, because now the formula had to be transposed. The latitude of the site was known (19º. 5), and the azimuth of the sunset was known (285º.5); what they had to solve for was the declination of the sun (in other words, its latitude) to yield the days on which such an event occurred. A few more moments went by and a series of hands went up. The answer: "The sun would be overhead at latitude 14º.8 North." I smiled and asked if that meant anything to anybody, and all the students shook their heads. I then suggested that they check the copy of the solar ephemeris we had taken along to see on which days the sun was overhead at that latitude. A brief examination of the ephemeris revealed that the sun passed through the latitude of 14º.8 N on April 30 on its way northward and again on August 13 on its way southward. "Does either of those dates mean anything to you?" I asked, and again I drew a total blank from the students. "That happens to be the latitude of Izapa, in southernmost Mexico," I explained, "and August 13 happens to be the date that the Maya believed the world began -- and here we find that date commemorated in the layout of the largest pre-Columbian city ever to be constructed in the New World, a thousand kilometers and a thousand years away from where it all began." What greater proof or sweeter vindication could any hypothesis ever receive? That day I felt as though I had been admitted to the inner sanctum of the Olmec priesthood, for I was privy to something which they knew and I knew, but no one had known for 2000 years in between.

Figure 27.

A view of Teotihuacán, the greatest metropolis of pre-Columbian Mesoamerica, looking southward from the Pyramid of the Moon toward the Pyramid of the Sun and the "Street of the Dead." Although a variety of reasons have been advanced for the city having been sited exactly where it is -- that is, in a somewhat offside valley about 15 km (9 mi) away from the former shoreline of the Lake of the Moon -- one intriguing clue may be that it lies directly in line with the winter solstice sunrise over Citlaltépetl (Orizaba), the highest mountain in all of Mesoamerica. 

               (This discovery was first published in 1978 and reconfirmed by two physicists at M.I.T. in 1980. At the time it was made, many of the details described in the preceding pages had yet to be filled in, so the entire story did not unfold as painlessly as might be surmised from the account given here. It is likewise interesting that nowhere in the literature is mention made of the fact that the positioning of the Pyramid of the Moon relative to the Pyramid of the Sun was carried out with such exactitude that for an observer standing on the former, the latter serves to mark the meridian -- i.e., a true north-south line. Thus, not only can the precise time of local noon be calibrated by the passage of the sun over the Pyramid of the Sun, but the zenithal positions of the moon and stars can easily be defined as well.)

Figure 28.

The largest metropolis of the pre-Columbian New World, Teotihuacán numbered in its heyday perhaps as many as 200,000 inhabitants. It is meticulously laid out on a grid which is offset 15º.5 from the cardinal points. Thus, its main avenue, the "Street of the Dead," runs from 15º.5 east of north to 15º.5 west of south, while its most impressive structure, the Pyramid of the Sun, is directly oriented to a point 15º.5 north of west -- the position at which the sun sets on August 13. The siting of the Pyramid of the Moon at the far end of the avenue was likewise done with such care that a sight-line directly over the top of the Pyramid of the Sun marks the meridian, thus allowing the priests of the city to fix the times of noon and midnight with complete accuracy.

               In the same 1978 article I made an almost heretical suggestion: I enclosed two maps of Mesoamerica which showed how more than two dozen key ceremonial centers in the region had been solsticially oriented toward the highest mountain within view, in some cases to sunrises, in others to sunsets, and in some instances to the winter solstice while in others to the summer solstice. On the map showing the Valley of Mexico and Teotihuacán I projected a line between the largest metropolis ever built in pre-Columbian Mesoamerica and the region's highest mountain, the 5700-m (18,700-ft) volcano, Orizaba, or Citlaltépetl ("the mountain of the star"), as it was known to the Aztecs. I had discovered this relationship -- a sunrise orientation at the winter solstice -- from my study of large-scale maps of the region, but was perfectly aware that Orizaba was not visible from Teotihuacán because of an intervening mountain ridge. Because the alignment hardly seemed to be a chance one, I hypothesized that there may have been some kind of "relay station" on the intervening ridge from which the priests, or their agents, might have been able to calibrate the winter solstice sunrise for the benefit of the residents of Teotihuacán, and I hoped at some time to be able to check out this theory in the field. From bearings taken from the Pyramid of the Sun and from computer maps supplied by CIMMYT, the Center for the Improvement of Maize and Wheat at Texcoco, and with the aid of a Geographical Positioning System receiver provided by the Trimble Navigation Corporation, I was able to locate the supposed site of this "relay station" with pinpoint accuracy during the field season of 1993. Interestingly, I found that a boundary or triangulation marker had been recently erected there (it had a date of 1991, apparently scratched into the cement while it was still wet) -- either to define a property line, a municipal limit, or a survey point -- and that it was surrounded by several irregular piles of stones, seemingly collected from the adjacent mountain top. Strewn around the crest of the hill, but heavily concentrated on the western side facing Teotihuacán, were numerous pottery shards representing many different types and styles of ceramic vessels, as well as several obsidian blades, most of which appear to have been used to line the edge of wooden swords (macana). Although these artifacts all date to Aztec times, the fact that many of them represent parts of braziers and bowls strongly suggests that the site was used as a religious shrine. (Deborah Nichols, personal communication). While the evidence is circumstantial at best that the site may have functioned as the author had originally proposed some 15 years earlier, its religious significance seems not to have gone unappreciated by Mesoamerica's last civilization (see Figure 23).

               Though Teotihuacán ultimately became the largest urban metropolis in Mesoamerica, it did not constitute the ultimate outlier of civilization on the edge of the great northern desert, or "la gran Chichimeca," as the Spanish later came to call it. At the time of the city's founding, the climate appears to have been moister over the Mexican plateau than it is today, and the frontier of Mesoamerican religion, commerce, and urbanism extended farther out into the basins and foothills of northern and western Mexico than it would have under the conditions which have prevailed over those regions in the last thousand years. Indeed, what surely must have been Mesoamerica's most forward outpost of urban settlement during this favored period of more adequate rainfall was the site now known as La Quemada, or Chicomostoc, in the central part of the state of Zacatecas.

               Strategically located near the headwaters of the Río Juchipila, the major northern tributary of the Río Grande de Santiago, Chicomostoc appears to date to the first centuries of the Christian era. Located on the top and flank of a commanding hill with a wide view over the northern approaches to the Juchipila Valley, it was clearly a fortified point erected as an advance outpost of central Mexican civilization against the nomad warriors of the north. The hill itself is of geologic interest, because it consists of a lava flow which has overridden an area of limestone. The latter stone breaks easily into tabular blocks, and it is these which were used to erect the great wall that circles the perimeter of the hill; the site's single, steep-sided pyramid; the imposing ball court which stretches out beneath it; and, most impressively, the great enclosed courtyard or palace with its huge, unmortared columns -- the so-called "Hall of Columns." The doorway of this structure, interestingly enough, looks out toward the western horizon at the sunset on August 13. Although the site's founding would appear to have closely followed that of Teotihuacán, its architecture seems to owe little inspiration to the great metropolis; in fact, it may well have been, in part at least, rebuilt by the later Toltecs, for multiple columns were one of their stocks in trade. Punctuating the limestone formation along the western and northern base of the hill are a series of caves which gave the site its indigenous name (Chicomostoc means "seven caves" in Nahuatl) and which also figure prominently in the legendary origins of the Aztecs.

Figure 29.

At Edzná the commanding structure known as Cinco Pisos has a view out across the flat expanses of the Yucatán Peninsula for 20 - 30 km (12 - 20 mi). At the base of the stairway leading up its western front stands an ingenious gnomon consisting of a tapered shaft of stone surmounted by a stone disc whose diameter is the same as the base of the shaft. At noon on the days of the sun's zenithal passage, the entire shaft is darkened by the shadow of the disc above it.


               Interestingly, about 1000 km (600 mi) away to the east a similar expansion of settlement was taking place about the same time. Sometime during the second century B.C. civilization was beginning its advance into the Yucatán region, carried by a people whom the archaeologists have called the "Pioneers." Surely, people of Mayan tongue had inhabited the region for many centuries in the past, but now for the first time the rise of a more advanced culture could be detected in the development of ceremonial centers with monumental structures. Although the Yucatán was a difficult environment in which to nurture an urban settlement, due to the lack of surface water, the undependable rainfall, and the stony character of the soils, the first major agglomeration in the region seems to have arisen in what has to have been the most favorably endowed area of the entire peninsula. At the risk of appearing to be a geographic determinist, it could be said that the Mayas had located the best place first!

Figure 30.

By the beginning of the Christian era, additional ceremonial centers had appeared in the highlands of Oaxaca and on the Mexican plateau. Among the Maya the earliest urban places had only recently been founded at Edzná and El Mirador.


               This was Edzná, in the interior of Campeche state about 50 km (30 mi) to the southeast of the port city of Campeche. It is located on the edge of what is the largest aguada or alluvial basin, in all of the Yucatán, and consequently had one of the most extensive agricultural hinterlands of any settlement on the peninsula as well as one of the more dependable water supplies. Excavations undertaken by Dr. Ray Matheny of Brigham Young University in the late 1970's revealed not only that Edzná was the earliest major Maya urban center, dating to about 150 B.C., but that at its peak it probably numbered some 20,000 inhabitants. Centered on a five story pyramid which the archaeologists have termed simply "Cinco Pisos" (Spanish for "five stories"), the site is laced by a series of great radial canals, first seen by Matheny in 1973 as he flew over the area near the end of the rainy season. Fully as impressive as the pyramids at Teotihuacán in terms of the man-days required for their construction, these canals were originally thought by Matheny to have some kind of astronomical orientation, but an astronomer he called in for assistance came up with nothing of consequence. When I first visited the site in 1978, it was with a different question in mind -- not why was the ceremonial center where it was, because that was already apparent from its siting on the edge of the aguada, but rather what orientations, if any, might the place reveal in terms of its internal structure. Cinco Pisos, I found, looked westward out across a large plaza which was bounded on its far side by a long ridge-like structure with a notch in its middle; and rising up behind the notch, was the top of a second, smaller pyramid. From the doorway of Cinco Piso's courtyard, through the notch, to the top of the small pyramid seemed to offer as carefully constructed a line of sight as could possibly be imagined, and when I measured its azimuth, I found that it was 285º.5 -- once again the sunset position on August 13! Indeed, checking the map of the site's layout as a whole, it was perfectly obvious that this orientation represented the guiding principle for all of Edzná, just as it had at Teotihuacán. Thus, in the Yucatán where the very flatness of the terrain all but precluded solsticial orientation to a mountain, the "summer solstice + 52 days" formula was already in use from the time of the first appearance of urban agglomerations in the landscape.

(Return to Table of Contents) (Continue to Chapter 6)