Mitla

← Encyclopedia

The floor is older than the church. Walk into the atrium of San Pablo Apóstol, in the town of San Pablo Villa de Mitla at the head of the Tlacolula valley, and the flat ground underfoot — the paved court where the parish gathers — is not the friars’ work but a pre-Hispanic platform they raised the church upon, the leveled stone of a sanctuary that stood here long before the first Mass was sung over it. The Cloud People who built that platform always said a door lay directly beneath: a stone door onto the cavern of the dead, which the Spanish missionaries, by their own account, filled in and sealed to stop the people returning to it. For four centuries that was a thing the town remembered and the record half-doubted. Then, in 2022, geophysical instruments passed over the church floor and found, some six meters down, a void in exactly the place the tradition had always fixed the entrance. The church has been standing on its own sealed underworld the whole time.

The two names

The place carries two names, and the doubling is the whole of its history in miniature. To the Zapotec — the Be’ena’a, the Binniza, the Cloud People of the Oaxaca valley — it is Lyobaa, a word the colonial sources gloss as “place of rest,” which is to say the place of tombs, the ground where the ancestors are laid down. When the Nahua of the Aztec sphere reached into the Tlacolula arm of the valley they translated that sense into their own cosmos and called it Mictlán, “place of the dead,” the deep ninth level of their underworld. From Mictlán, worn down through Spanish ears, comes the modern Mitla. The name the whole world now uses for the Zapotec necropolis is a foreign translation of its meaning, laid over the Zapotec word like the church over the platform.

The site sits at the upper, eastern end of the Tlacolula valley, about 44 kilometers southeast of the city of Oaxaca, at roughly 1,480 meters — some 4,855 feet — above the sea, in the modern municipality of San Pablo Villa de Mitla. Its coordinates are 16.92722° N, 96.35944° W. This is dry, warm country at the valley’s throat, where the fertile floor narrows toward the mountains, and it was inhabited early: the Wikipedia synthesis of the archaeology puts continuous occupation from at least the Classic period, roughly 100 to 650 CE, with signs of settlement reaching back perhaps as far as 900 BCE, and the great stone palaces for which the place is famous belong to the centuries of its peak, between about 750 and the Spanish arrival in 1521. In those late centuries the architecture takes on a strong Mixtec–Zapotec character, the two neighboring valley traditions meeting in the same walls; the Mixtec presence in the eastern arm of the valley grew heavy in the Postclassic, and Mitla’s mature style is usually read as the product of that meeting rather than of either people alone.

The record of the place comes in two very different layers, and they answer two different questions. The stone itself — the platforms, the mosaic palaces, the tombs cut beneath their floors — is the primary witness to what was built, and it can be dated and measured and mapped. But the stone does not say what the buildings meant to the living Zapotec who used them, because the city that used them was still in use when the Spanish came, and its meaning had to be written down by outsiders. That second witness is the colonial texts: Fray Toribio de Benavente, called Motolinía, who noted the fret-worked buildings in the early-to-mid sixteenth century; Diego García de Palacio, whose account is dated 1576; and, fullest of all, the Dominican Fray Francisco de Burgoa, whose Geográfica descripción of 1674 gives the first extended description of what lay at Mitla and beneath it. These men are indispensable and they are unreliable in a precise way: they are the primary source for the Zapotec meaning of the place, and they came to end what they were describing. Their pages have to be read against their own grain — the friars’ vocabulary of idols and devils set aside as the record of a persecution rather than a description of a religion — but read that way, they are nearly all the interior testimony there is.

A script without mortar

What Motolinía and the rest saw first, and could not stop describing, was the fretwork. The palaces of Mitla are sheathed in grecas — panels of stepped-fret and diamond mosaic, the stone laid in the interlocking geometric key-patterns that give the site its whole visual character. The technique is what astonishes. Each panel is built up from thousands of small stones, individually cut and polished and fitted together so tightly, and set into the wall so exactly, that they hold in place without mortar — a veneer of dry-laid mosaic, sometimes over a red-stuccoed core, running in long registers along the palace facades and around the interior courts. There are said to be more than a dozen distinct fret motifs across the complex, and the detail most often repeated about them is that none is repeated: as the record of the site has it, no fretwork design is duplicated exactly anywhere in the whole of Mitla. Whether that is literally true down to the last panel is hard to test, but the impression it names is real — a vast decorative program that never resolves into a single stamped pattern, a script of order in stone that never quite repeats its own word.

The grecas do not stand alone; they clothe a planned complex. The archaeological zone is conventionally divided into five architectural groups: the South Group, the Adobe Group, the Arroyo Group, the Columns or Palace Group, and the Church or North Group. The two best preserved are the Columns and the Church groups, and it is the Columns Group that holds the most famous building, the long Hall of Columns — a rectangular room whose flat roof was once carried on a row of massive monolithic stone pillars, opening onto an inner patio walled on every side with the mosaic panels. These are not temples in the pyramidal Mesoamerican sense; they are low, horizontal, palatial buildings set around courtyards, an architecture of enclosed patios and long decorated halls rather than of stairways climbing to a summit. The whole vocabulary is different from the great vertical capital of the earlier Zapotec state, Monte Albán, forty-odd kilometers to the west, which had already fallen into ruin centuries before Mitla’s palaces rose. Mitla is the valley’s later thing: not the mountain-city of the living state but the ground-level house of its dead.

The Hall of Columns, between the courtyards of the Columns (Palace) Group at Mitla — a long flat-roofed hall once carried on a row of monolithic stone pillars. Photograph by DavidConFran, via Wikimedia Commons. CC BY-SA 3.0.

The house of the dead

Within the Zapotec sacred world Mitla is a threshold — the door between the country of the living and the country of the dead, and so the necropolis and the oracle-seat of the later Zapotec order. Its patron is the lord of the underworld himself, named in the colonial sources as Coqui Bezelao, also rendered Pitao Bezelao, the Zapotec keeper of the dead, to whom the great dead of the valley were brought; the fuller account of that power and his consort belongs with the entry on Coqui Bezelao. It is worth marking at once that his identity is one of the contested registers here. The name comes to us through the late and colonial record — the friars’ chronicles and the sixteenth-century Zapotec vocabulary of Juan de Córdova — and reconstructing a pre-conquest theology from post-conquest pages is genuinely interpretive work, the sort that scholars from Alfonso Caso to Joyce Marcus have warned must be handled as inference rather than as settled fact. What is not in doubt is that Mitla was a place given over to the dead.

At its head stood the high priest, the Uija-tào — the name is usually glossed as “the Great Seer” — who resided at Mitla and whose spiritual authority rivaled and at times exceeded that of the secular Zapotec lords. The Spanish, groping for a comparison their readers would grasp, likened him to a pope. He was consulted before war and before the succession of kings; he presided over the rites for the dead; and he was still resident at Mitla, keeping the door, when the Spanish reached the valley in the 1520s. Beneath the palace floors lay the tombs he tended. Cut into the ground under the buildings are cruciform tombs — chambers shaped as a cross, entered by a stair that descends from the patio into the earth, their interior walls lined with the same greca mosaic that runs along the walls above. Into these the kings and the priests were laid. The living and the dead of the Zapotec elite were stacked in the same architecture: the fret-worked palace above, the fret-worked crypt below, the one a house and the other a grave, decorated alike.

Burgoa’s 1674 chronicle carries the Zapotec account of what lay deeper still, and this is where the record thins into tradition. His informants held that beneath the priests’ residence ran a set of interconnected subterranean chambers — as Burgoa reported it, a temple of four rooms, the first a sanctuary, the second the burial vault of the high priests, the third the tomb of the kings of Teozapotlán, the seat the Spanish knew as Zaachila. From the last chamber, they said, a stone door opened onto a deep cavern that ran on underground — thirty leagues, by the figure Burgoa records, a labyrinth of passages held up by stone columns. That cavern was the mouth of Lyobaa itself, the subterranean palace of the dead, into which the honored dead and, it was said, the world-weary living might be received. Here the registers must be kept apart with care. The chamber-count and the thirty leagues are a reconstruction from a thin, late, second-hand source and cannot be taken as survey data; the door onto the cavern as the living entrance to the country of the dead is tradition-internal meaning, the sense the place carried for the Cloud People, which is a different kind of truth from a measured fact and is stated here as theirs.

Over all of it the greater gods of the valley presided from a distance. Cocijo, the lightning-lord of the four quarters whose fanged mask stares from the gray funerary urns, is the storm the valley’s harvest depends on; Pitao Cozobi is the maize made a god. But rain and maize belong to the sky and the fields. Mitla is given over to what lies under them — to the ancestors and to the earth that takes them in — and the storm-god’s mask, common enough on the urns set with the dead across the valley, keeps its watch at the threshold of a house that is not, finally, his.

A detail of the grecas on a wall of the Hall of Columns, thousands of small cut stones fitted without mortar into stepped-fret mosaic, traces of the old red stucco still visible. Photograph by DavidConFran, via Wikimedia Commons. CC BY-SA 3.0.

The sealed door and the church

The frontier at Mitla is a religious persecution, and the persecution is told in stone. In 1553 the ancient sanctuary was ordered broken — the Oaxacan record lays the order to the church authorities of the diocese, naming the prelate Albuquerque — and the order was carried out: several of the mosaic palaces were pulled down and quarried for their cut stone, and the colonial Church of San Pablo Apóstol was raised in the North group directly upon the pre-Hispanic platform, which still serves as its atrium to this day, its very fabric partly the dressed stones of the older sacred complex. The church did not merely stand near the sanctuary. It stood on top of it, sited so that the geography of the saved was built directly over the geography of the dead.

Burgoa, writing in 1674, gives the fullest early account precisely because he stands at the moment of extirpation and records what was being undone. He describes the subterranean temple, the four chambers, the stone door onto the cavern — and then records that the missionaries, appalled at what they took for a sanctuary of the devil, had all the entrances to the labyrinth sealed and filled, to stop the Zapotec from going back down into it. That sealing is the load-bearing fact of the reception, and the frame around it is the thing to name and refuse. The friars called what they walled up idolatry, superstition, devil-worship; that vocabulary is a historiographic problem, not a description to be adopted. What they filled in was not superstition. It was the funerary heart of a literate valley civilization — the crypts of its kings and the seat of its high priest — stopped up with rubble so that a people would stop mourning where they had always mourned. The wider campaign against Zapotec religion, the trials and the burning of the calendar books that ran on into the eighteenth century, belongs with the history of the Zapotec as a people; at Mitla its monument is simply this, a church standing on a filled door.

The stone outlasted the sealing. In the twentieth century Alfonso Caso and the archaeologists who followed him cleared and studied the five groups, and Mitla took its place among the recognized monuments of Mexico, drawn into the heritage sphere of the Oaxaca valley and its World Heritage inscriptions; today it is, as the record notes, the second most visited archaeological site in the state of Oaxaca after Monte Albán, its fret-worked walls photographed by the busload and sold in miniature at the gates. The commerce is part of the site’s modern life and part of what has to be seen clearly about it: the grecas that sheathed the crypts of the Zapotec dead are now a design motif, reproduced on tiles and textiles and postcards, the funerary architecture of the Cloud People turned into the decorative signature of a tourist valley.

And the door itself has been found again, not by priests but by geophysicists. In 2022 the Project Lyobaa survey — a collaboration of Mexico’s National Institute of Anthropology and History, the National Autonomous University of Mexico, and the ARX Project — ran three kinds of non-invasive instrument across the Church group: ground-penetrating radar, electrical resistivity tomography, and seismic noise tomography, combined for the first time in Mexico into a single three-dimensional image of what lay beneath. The scans found, directly under the Church of San Pablo, a substantial void about six meters down, opening toward deeper and still-unexplored cavities — a network of subterranean chambers and passages in exactly the place where Burgoa’s 1674 chronicle and unbroken local memory had always put the entrance to the underworld temple. The result wants stating carefully. It is a documented detection of chambers, not a proof of a mythic cavern; the radar found a void where the tradition said a door was, which is a remarkable convergence and not a descent into Lyobaa. The account nearly three and a half centuries old turned out to describe something that is really down there. For now the mouth of Lyobaa is mapped rather than opened, known by its echo, sealed still.

The Be’ena’a of San Pablo Villa de Mitla live around it. The town is not a dead ruin with a parish attached but an inhabited Zapotec place, its language spoken, its dead still carried the short distance to the ground that has always received them, its festivals for the returning dead kept year on year above the filled crypts. The friars stopped the door and set their church on its lid; the radar found it anyway, six meters down under the altar floor; and in the streets that ring the churchyard the Cloud People go on making their way home to the ground beneath it, which is the one arrangement at Mitla that four centuries above it never managed to change.

→ Related: Cocijo · Pitao Cozobi · Coqui Bezelao · Dzahui · Nine Wind · Monte Alban · Apoala · Tilantongo · The Zapotec · The Mixtec · The Mixe