The Sun Temple Of Colorado's Mesa Verde Just Got Even Stranger

On a juniper-covered mesa in Colorado's far southwest corner, a canyon's width from a city built inside a cliff sits a giant letter D made of stone. From the air, it looks like a seal pressed into the earth. Its walls run roughly a thousand feet, about 28,000 cubic feet of masonry on a footprint bigger than a basketball court. Around 1250, you would not have heard wind in the junipers. You would have heard a region of roughly 19,000 people. Today, nothing. Archaeologists call it Sun Temple, and almost nothing about it behaves like a building. Its outer wall has no doorway, no roof beams, no hearths, no trash. At the winter solstice, the setting sun drops between its towers. And then came the detail that made no sense. The sun symbol that gave this place its name was never carved by human hands. The accepted answer sounded simple. A temple left half-built when its makers fled. But nobody raises that much stone and forgets the door. Into that gap rushed the bigger theories. A solar observatory, a lunar calendar, the golden ratio. For a century, this mystery was argued with shovels and guesses. Then, in 2016, a statistician at Arizona State University tested its wall lines against the sky using satellite imagery and probability math. What came back did not end the argument. It cracked it wider open. So what were the people of this mesa really building? And why did they walk away? Look at it from the air, and the shape hits you first. A perfect capital D on top of Chapin Mesa in Mesa Verde National Park. The straight edge runs along the south. The curved wall sweeps around the north like the back of a shield. No sprawl, no village growing messily around it. Just one clean, deliberate, geometric statement staring across the canyon at one of the most famous ruins in North America. That ruin is Cliff Palace. Roughly 150 rooms tucked under an overhang. Towers and plazas and packed into the rock like a city hiding from the sky. Mesa Verde became a national park in 1906, the first one in the United States created to protect the works of human beings rather than mountains or geysers. In 1978, it was named one of the very first world heritage sites on Earth. This is not some forgotten pile of rubble. This is a crown jewel, and Sun Temple sits directly across from it, facing it, as if the two were meant to be seen together. So what does the official story say? The National Park Service keeps it careful. Pueblo people, the descendants of the people who built this place, classify it as ceremonial. Some believe it was never completed. The walls once stood somewhere between 11 and 14 feet high, and a small basin carved at one corner may have served as a kind of sundial. It sounds calm, reasonable, settled. It is not settled. These are double walls, built core and veneer, meaning two faces of carefully shaped sandstone with a packed rubble core filling the gap between them. That is how you build when you want something to last. Inside the D, archaeologists counted roughly two dozen rooms of different shapes and sizes, plus three circular chambers built into the plan. A fourth circular structure stands detached outside to the southeast. Rooms within rooms, walls within walls. Now walk the full curve of that D, then the long straight wall on the south. You will not find a ground-level entrance cut through it. No floors covered with the usual debris of daily life, no piles of broken pottery and bone that archaeologists find around almost every ancestral Pueblo home. When the building was dug out, it was almost eerily clean. Think about the labor. Thousands of stones quarried, shaped, carried, and fitted by people without metal tools, without draft animals, without wheels. A whole community spent seasons of its life on this. And then, apparently, nobody lived in it. Nobody roofed it. Nobody filled it with the mess of being human. So why build it? Every confident answer to that question has ended up cracked. The date was wrong. The alignment was wrong. The famous symbol at its corner was wrong. Even the idea that it was unfinished is now under attack. To understand why, you have to go back to the man who named it, and to the one strange stone that convinced him he had found a Temple of the Sun. Because the name itself was a mistake. Go back to December 1888. Snow on the mesas. Two ranchers, Richard Wetherill and his brother-in-law, Charlie Mason, are riding the canyon rims looking for stray cattle. They look across a gorge and see it. An entire stone city hanging inside the cliff. They name it Cliff Palace. Within a few years, diggers and collectors are hauling pots, baskets, and human remains out of the canyons. In 1891, a young Swedish researcher named Gustav Nordenskog arrives. He is careful by the standards of his time. Measuring, photographing, documenting. He is also digging. And when he leaves, a large collection from Mesa Verde leaves with him, bound for Europe. Hold on to that. It comes back at the end of this story, and it changes the meaning of everything. After the park is created, the Smithsonian sends in Jesse Walter Fuchs, an ethnologist who spends years repairing and interpreting the cliff dwellings for a public just starting to arrive by road. And from there, Fuchs keeps noticing something on the opposite mesa top. A long, low mound covered in soil and brush, shaped like nothing much at all. In 1915, he decides to dig it. And here is the first suspicious detail, written in his own report. Fuchs explains that he chose this mound partly because tourists visiting Cliff Palace could see it. Read that again. The site was selected not only for its scientific promise, but for its view. A new park needed attractions. And the man in charge of uncovering this mound was also the man shaping what visitors would be told about it. That does not make Fuchs a fraud. He was a respected scientist working in a new field with almost no tools. But it does mean the mound was being turned into a story before the shovels even went in. There is a second odd detail. In his report, Fuchs writes that Nordenskold, two decades earlier, never mentioned this mound, as if it had been overlooked, untouched, waiting for him. But when later researchers went back through the record, they found an 1891 Nordenskold photograph showing the mound. The Swede had seen it, possibly photographed it on purpose. The clean narrative of a hidden site, discovered at last, was already a little too clean. Then the digging begins, and the D rises out of the dirt. In the old Smithsonian photographs, a shapeless hump becomes walls, then corners, then the full curve of the D. Later, an automobile drives slowly around the excavated ruin, a sign of exactly who this site was going to serve. The motor age had reached the mesa, and the ancient building was ready for its audience. But Fuchs had a problem. He needed to say how old it was. This was before tree ring dating could give exact calendar years. So Fuchs improvised. Growing on top of the mound was a cedar. He cut it, counted its rings, and got about 360 years. Then he added what he himself called a conservative guess of 250 years, for the time it must have taken for the ruin to collapse and be buried before the tree took root. His arithmetic pushed the building back to roughly the year 1000. It sounded scientific. It was a stump and a guess. In 1929, the astronomer A.E. Douglas and his colleagues finished building a continuous tree ring calendar for the southwest. Suddenly, beams pulled from Mesa Verde's cliff dwellings could be dated to the exact year their trees were cut. And the dates did not cluster around 1000. They clustered in the 1200s. Now here is the twist you do not see coming. Tree ring dating needs wood. Roof beams, lintels, posts. And Sun Temple has no roof. No beams. Almost no timber at all. Which means the most powerful dating tool ever used in the southwest has almost nothing to grab onto here. The dates that you usually hear, somewhere in the mid to late 1200s, are inferences. Educated guesses built from masonry style and from the dates of the buildings around it. The building's emptiness does not just hide its purpose. It hides its age. But the date was not the strangest thing Fuchs found. The strangest thing was at the southwest corner. Built into the base of the wall there, Fuchs found a stone with a peculiar marking. A rounded depression with lines radiating out from it, like rays. To Fuchs, it looked like a sun. And around it, the builders had constructed a small enclosed space. Almost like a box built to protect the stone. Fuchs was electrified. He called it the Sun Shrine. He suggested it may have been connected to the sunset at the September Equinox. And from that single stone came the name that has stuck for over a century. Sun Temple. The doorless walls now fit a story too. Fuchs argued the building had been designed for secrecy. For He considered whether it might have been a fort and rejected the idea, pointing out that the walls had no portholes for defense. He noticed it looked a little like the famous D-shaped great house of Pueblo Bonito in Chaco Canyon, then dismissed the resemblance as superficial. Everything was clicking into place around one idea. A sacred building for the sun, with a sacred symbol at its corner. And then Fuchs did something that would horrify archaeologists today. To preserve the ruin for visitors, he capped the tops of the walls with Portland cement. Every wall. He also added masonry to level the outer walls, evening out what time and collapse had left behind. Elizabeth Dickey, who leads cultural resources at the park, has said that preservation practices like these would leave us aghast today. Because when you add stone to an ancient wall, you change the evidence. You make it harder for anyone after you to know exactly where the original builders stopped. Remember that, because the biggest question about Sun Temple is precisely where its builders stopped. But first, the stone. The sun. The thing the entire name rests on. Researchers were already uneasy about it. The marking had been described in some accounts as a fossil palm leaf, an identification made from a poor photograph. Then the geologist Willis T. Lee looked at it and offered a different idea, calling it a feature left by an ancient spring from the age of the dinosaurs. A sun symbol. A fossil leaf. A Cretaceous spring. Three explanations for one rock. Something was wrong. In 1932, a man named Walter Lang studied it closely, and the park's own publication, Mesa Verde Notes, reported his conclusion in July 1933. It was an erosion basin. A natural pattern made by water and weather wearing away at soft sandstone. Not carved. Not painted. Not a fossil. And to make the point unmistakable, about 20 similar marks were identified elsewhere in the park. Let that land. The rayed sun that gave Sun Temple its name was weathering. The same kind of weathering found on stones all over Mesa Verde. Now, that does not mean the builders ignored it. They did build a special enclosure around that stone. Maybe they saw something meaningful in the pattern, exactly as Fuchs did. But it means the most famous piece of evidence was not a symbol made by human hands. And if the symbol was wrong, what about the sun? That is when the measuring started. In 1977, the archaeologist Jonathan Raymond published a paper with a title that does not hide its purpose. Solstice misalignment at Sun Temple, correcting Fuchs. Fuchs had suggested the long south wall of the D lined up with the solstices, the turning points of the solar year. Raymond measured it, and the wall did not line up with the solstice sunrise or sunset. It did not line up with the equinoxes either. It was off. That should have ended the solar theory. It didn't. Because the people who built this region were obsessed with the sky. That is not speculation. Pueblo communities today still keep ceremonial calendars tied to the sun's movement along the horizon, with sun watchers tracking where it rises and sets against distant landmarks. Across the ancient southwest, researchers have documented light and shadow effects on rock art and architecture. At Hovenweep, just to the west, the National Park Service describes Hovenweep Castle, with two D-shaped towers, where one beam was cut in 1277, and where openings let sunlight mark the solstices and the equinoxes. The idea that these builders encoded the sky into stone is not fantasy. So maybe the question was wrong. Maybe the sun was not lined up with a wall. Maybe it was lined up with something else entirely. Enter the astronomer J. McKim Malville and the researcher Greg Munson. Instead of standing inside the D and looking out, they asked what Sun Temple looked like from across the canyon, from Cliff Palace, from the place people were actually living. They brought precision surveying instruments. They mapped horizons. And in December 1997, on the winter solstice, Munson stood at an observation point near Cliff Palace and watched the sun go down. As described by the High Altitude Observatory in Boulder, at winter solstice, seen from the observing station, the setting sun touches the horizon between Sun Temple's two main towers. Picture it. The shortest day of the year. The sun sinking lower and lower, sliding along the horizon day after day, reaching its southern limit. And from the edge of the cliff city, it drops precisely into the notch formed by the mysterious building across the gorge. And there was more. The smallest of the towers, according to that same description, might have served as a horizon marker to anticipate the solstice by about 20 days. A warning. A countdown. Get ready. The sun is about to turn around. So the solar theory came back. Not in the wall Fuchs measured, but in the view from the other side of the canyon. And then the moon entered the story. The sun's path along the horizon repeats every year. The moon is far stranger. Its rising and setting points swing back and forth on a cycle that takes about 18.6 years to complete. At the extreme of that cycle, called a major lunar standstill, the moon rises and sets farther north and farther south than at any other time. It happens roughly once in a generation. You can live your whole adult life and only see it two or three times. In June 2006, near a major standstill, Munson's team watched from the painted tower in Cliff Palace as the moon set at its southern extreme. They watched again on June 2nd, 2007. And both times, the moon went down over Sun Temple. Now look at the timing. Munson and his colleagues linked this alignment to the major standstill of 1279 to 1281. Those years fall right in the last decades of the people who built Cliff Palace and Sun Temple. A once-in-a-generation lunar event, possibly framed by the building across the canyon, at almost exactly the moment this entire world was coming apart. One theory from Munson's team is that Sun Temple was a gathering place, a monument where people from across the area came together at moments marked by the sun and by that rare long cycle of the moon. Not a house, not a fort, a clock for a community. But they went further, and this is where the old story really breaks. In 2006, Munson, together with Larry Nordby and Douglas Bates, carried out detailed architectural documentation of the building. Not glancing at it, reading it, stone by stone, joint by joint, the way you would read a They studied how walls met, which ones abutted which, where one phase of building ended and another began. And they reconstructed the sequence. The circular structures came first, Kiva's A, B, and C, and they appear to have started life as towers. Then the great D-shaped wall was built around them. Then later, an annex was added. It was not one rushed project. It was building that grew in phases, with planning and intent. And then they wrote a sentence that flips the whole mystery on its head. We currently believe that no portion of Sun Temple was roofed or was ever designed to be roofed. Sit with that. For a century, the story was simple. The people of Mesa Verde started an ambitious temple, ran out of time as their world collapsed and walked away from it half built. Missing roof, missing doors, missing finish, a monument to interruption. But what if nothing is missing? What if the builders made exactly what they wanted? An open air enclosure, walled off from the ground, open to the sky, built precisely so that the sky was the point. A building where the roof was never part of the plan because the ceiling was the sun, the moon, and the stars. That does not prove the wildest version of the story, but it does make the simple version, the unfinished temple, much harder to defend. And remember Fuchs and his cement. When a building has been leveled and capped by an excavator, telling the difference between a wall that was abandoned halfway and a wall that was meant to stop there gets very, very difficult. The unfinished building might be finished. The doorless walls might be deliberate. And the mystery is no longer what went wrong. The mystery is what they were trying to do. That is where an Arizona State University researcher named Sherry Towers pushed the story into even stranger territory. In 2016, Towers, a statistician by training, analyzed the site using satellite imagery and Google Earth. She identified 22 sight lines built into the structure, lines formed by walls, corners, and features. Then she asked how many of them pointed toward significant events in the sky, rising and setting points of the sun, the moon, and bright stars and star clusters tied to Pueblo tradition. Her answer? About 86% of them matched celestial bodies sacred to Pueblo people. And she put a number on how unlikely that was by chance, a p-value of 0.0023. In plain language, if you scattered random lines across a building, you would expect a result like this roughly once in about 400 tries. She found one more detail that is hard to shake. From the sun shrine, the box built around that weathered stone, a sight line points toward the rising of the Pleiades, the small star cluster that many cultures across the world use to mark seasons and planting time. Then in 2017, Towers published a second paper, this time on the geometry of the plan itself. She argued that the builders used a standard unit of length of about 30.5 centimeters, plus or minus half a centimeter. She found 3-4-5 right triangles worked into the layout, and she found rectangles in the plan whose proportions came close to the golden ratio, that famous number of roughly 1.618. Her measured ratios were 1.643 and 1.646. If she is right, the people who built Sun Temple were not just stacking stones. They were laying out a precise, mathematically designed ceremonial observatory in the 12th or 13th century without written numbers, without compasses, without anything we would recognize as a drafting table. And this is the moment where you have to slow down, because Towers herself said something remarkable. Talking about her work, she admitted that you can find alignments in almost any building if you go looking for them. That is the trap of archaeoastronomy. The sky is full of targets. Sun, moon, stars, rising points, setting points, solstices, equinoxes, standstills. The ground is full of lines. Every wall has two ends. Every corner points somewhere. With enough targets and enough lines, coincidence can start to look like design. That is why one statistical study is never the end of the argument. So what is the honest status? Towers' work is real, published, and statistically significant on its own terms. But it comes from a single author. It relies on remote imagery rather than full surveying on the ground. And as far as published work shows, nobody has independently replicated it. Look closer at those golden ratio numbers too. 1.643 and 1.646 are about one and a half percent higher than the golden ratio. Close, intriguing, but not exact. And there is a cautionary tale just down the road. Researcher Anna Sofer and her solstice project spent decades arguing that great houses in Chaco Canyon, including D-shaped Pueblo Bonito, were oriented to solar and lunar events, even to the lunar standstill. It became one of the most famous ideas in southwestern archaeology. Then researcher Andy Monroe surveyed 28 Chacoan sites with a compass and a theodolite. His conclusion was blunt. He found no convincing evidence for previously proposed architectural alignments to lunar standstills, June solstice sunrises, or equinox events. Same region, same kinds of buildings, same kinds of claims, and a careful survey that came back empty. So you are left holding two things at once. A winter solstice sunset that drops between the towers, seen by an actual human being standing at the cliff edge. A southern moon set over the same building, observed twice. A statistician saying the numbers are too good for chance. And a long history of alignment claims in this region that have collapsed under closer measurement. Every answer this building has given has turned out to be partly true and partly a projection. The sun symbol was weathering, but the builders still enclosed it. The south wall misses the solstice, but the view across the canyon catches it. The building looks unfinished, but it may have been complete. The geometry looks perfect, but it is just slightly off. The building does not lie, but it does not confess either. To get closer to the truth, you have to stop staring at the stones and start looking at what was happening to the people who stacked them. By the mid-1200s, the Mesa Verde region was crowded. The Village Ecodynamics Project, a long-running research effort that modeled settlement across the central Mesa Verde region, estimated a peak of around 19,500 people between 1225 and 1260. This was not a fragile outpost. It was one of the most densely settled landscapes in the ancient southwest. Then everything turned. Tree rings record a severe drought that the pioneer A.E. Douglas called the Great Drought, running from 1276 to 1299. Crops failed. Water sources shrank. But here is the detail people often miss. The exodus had already started. The village ecodynamics researchers concluded that regional abandonment probably started in the west in the late 1260s and in the central area in the 1270s and that it was complete by the early or middle 1280s. People were leaving even before the worst years of drought hit. And not everyone left peacefully. At a village called Castle Rock, not far from Mesa Verde, archaeologists found the remains of at least 41 people who died violently around 1280. At Sand Canyon Pueblo, at least eight more. The end of this world was not a quiet migration in the night. In places, it was a catastrophe. Then look at the building record. Across Mesa Verde, the tree ring dates that tell us when people were cutting timber for construction almost disappear. Only about five dates fall in the 1280s. After centuries of building, the hammering simply stops. And by around 1285, the region was essentially empty of the people who had lived there for generations. Now put Sun Temple back into that timeline. A massive, carefully phased construction project probably begun in the middle or later 1200s. A building that seems to frame the winter solstice and possibly the southern extreme of the moon. A lunar standstill falling from 1279 to 1281, right in the heart of the collapse. Drought. Violence. Departure. Was Sun Temple an attempt to hold a community together while it was falling apart? A place to gather, to mark time, to perform ceremonies at exactly the moment people needed the sky to promise that the world would turn again? Or was it a monument finished and then deliberately left behind as entire communities walk south and east toward new homes along the Rio Grande and in Arizona? We cannot prove either, but notice how the meaning flips depending on one detail. If it was unfinished, it is a tragedy frozen in stone. If it was finished, it is something much more haunting. A completed sacred place built for the sky that its makers chose to walk away from. And this is where the story stops being only about stones and stars. Because the people who left did not vanish. For a long time, the builders of Mesa Verde were called the Anasazi. The word comes from Navajo, and it is often glossed as Ancestors of Enemies, or Enemy Ancestors. Not a name the builders' own descendants chose. Today, the National Park Service uses the term Ancestral Pueblo People, and the park recognizes 26 affiliated tribes. The Hopi, the Zuni, Acoma, Zia, and the Rio Grande Pueblos are among the living communities who trace their history to these canyons. That changes the question, because the mystery of Sun Temple is not a mystery to everyone in the same way. Now remember the young Swede, Gustaf Nordenshold, and his collection. In 1891, his excavations removed more than 600 items from Mesa Verde, including human remains. They crossed the ocean. They ended up in the collections of the National Museum of Finland in Helsinki. For more than a century, people buried in these canyons rested on storage shelves thousands of miles from home, catalogued, labeled, studied. It took until 2019 for the process of bringing them back to gain real momentum, including discussion between the President of Finland and the President of the United States. And in September 2020, the National Museum of Finland returned the remains of 20 ancestors, along with 28 funerary objects. They were handed over to the Hopi, Acoma, Zia, and Zuni, and reburied at Mesa Verde in mid-September 2020. Think about what that means for a story about a building nobody can explain. The same era that dug out Sun Temple, capped its walls with cement, and gave it a misleading name, also carried away the dead who once walked these mesas. And the people with the deepest connection to it were, for a very long time, barely asked. When the Colorado Sun reported on the debate over Sun Temple's purpose, descendant voices offered something you will never find in a statistical table. Filip Tuoletstiwa, a Hopi geodesist, a scientist who spent his career measuring the shape of the earth, offered a different way to look at the site. Not as an instrument that measured cycles, but as a place where communities came together to celebrate them. And he pushed back on one of the most common words in this whole story. We don't like that word. Abandoned, he said. These are living, breathing places. And Elizabeth Dickey, the park's cultural resources leader, described what tribal members have told the park when asked to explain sites like this one. It's not for us to know. That sentence lands differently once you know everything else. Because maybe the most powerful reversal in this story is not about astronomy at all. For a century, outsiders treated Sun Temple like a locked box that could be forced open with the right tool, the right excavation, the right stump to count, the right wall to measure, the right satellite image, the right formula. And every time, the box gave back something partial. A weathered rock instead of a sun symbol. A misaligned wall instead of a solstice marker. A completed building instead of an abandoned one. An elegant statistic that its own author warned could be fooling us. Maybe some of the answer was never meant to be dug up. Maybe the gap is not a failure of science. Maybe it is a boundary. That does not mean the research is wrong to ask. The solstice sunset between the towers is real. The moonset over the D was observed. The careful architectural reading of how the walls were built is real work. The questions are worth asking. But the deeper question is who gets to hold the final answer? The researchers with their instruments and p-values? Or the descendants of the people who stacked those stones? Who still say this is a living place and who choose what stays unspoken? Now go back to the beginning. Back to the cliff city. Looking across the canyon at that stone D as the winter sun slides down between its towers. But this time, it is not the sun you are waiting for. The moon's long 18.6 year cycle reached its most recent major standstill in 2024 and 2025. For the first time in roughly a generation, the moon swung to its farthest southern setting point. If Munson and his colleagues are right, then once again, after more than seven centuries, the full moon slid down the sky towards Sun Temple. The way it may have done for the people of Cliff Palace in 1279, 1280, and 1281 in the very last years before they left. And now it is gone again. The next major standstill is not expected until around 2043. So picture it one more time. Night over Mesa Verde. The canyon black. The cliff city behind you silent. The moon sinking at the edge of its southern limit. Lower, lower, toward a roofless stone D that has no door, no beams, no trash, no confirmed date, and a name built on a weathered rock. Moonlight slides along the tops of its thick walls. And there, along every ridge, the gray line of Portland cement glints back. A 1915 seal laid across a 13th century question. The moon goes down over Sun Temple, and the building keeps its answer.