At an altitude of nearly two kilometers in the Norwegian mountains, a freezer no one ever designed is melting—one that worked flawlessly for several thousand years, until it broke down.
In 2006, after two snowy winters and a hot summer, systematic collection began on the slopes of the Jotunheimen range in central Norway of things that should not have existed. Woolen cloth, leather shoes, wooden shafts—organic matter that under normal conditions rots away within decades—lay on the surface of a retreating ice patch, as though someone had dropped it only moments ago. The program, called the Glacier Archaeology Program, was led by archaeologist Lars Pilø, field director Espen Finstad, and medieval economy specialist James Barrett. Their workplace was Lendbreen Pass: the ice patch begins there at 1690 meters, rises to 1920, while the pass itself sits at 1973 meters. As it retreated, the ice exposed a band of artifacts dozens of meters wide and several hundred meters long. The pass itself is steep, windswept—a passage between valleys where meltwater squelches underfoot in summer, and everything vanishes beneath snow again in winter. This was where the ice cover played a role no one fully suspected until the late 20th century: it was a preservative.
Ice is a storage environment with very specific properties. Below-zero temperatures suppress bacterial activity. The absence of liquid water halts hydrolysis—the chemical breakdown that destroys collagen in leather and cellulose in wood. The mechanical pressure of an ice cover, if it is stationary, holds an object in place like a paperweight. But the key word is “stationary.” As long as the ice remains still, an object can lie inside it for centuries. The moment it begins to melt, the countdown starts. Meltwater soaks into wool and wood, expanding in nighttime frosts and tearing the fibers apart from within. At an altitude of two kilometers, wind dries the surface out in hours. Microorganisms that have been waiting in the soil beneath the ice receive moisture and a nutrient substrate at once. One field season—and in the Norwegian mountains it lasts only a few weeks—can erase an object that remained intact for a thousand years.
At first, the finds were scattered: local herders and hunters picked up strange objects at the edge of melting ice and sometimes reported them to museums. But when archaeologists began working the pass systematically, a structure started to emerge. Stone cairns—hand-stacked pyramids of rock—marked the route. The remains of shelters built from flat slabs pointed to rest stops. Packhorse dung, preserved in frozen soil, spoke of regular caravan traffic. Horseshoes—dozens of horseshoes—confirmed that animals had been shod here, which meant they were valued enough for someone to care for their hooves before a difficult crossing. The pass was a working road—essentially a mountain highway, even without asphalt.
One detail gets me: alongside tools and gear lie things that were simply lost. A mitten. A lone shoe. A broken piece of a sled. This is not a hoard and not a burial—it is the trace of people crossing the pass on business, freezing, breaking their equipment, losing gloves, and moving on. That everyday noise of a route cannot be faked, and it is almost impossible to find in ordinary excavations, where organic matter vanished long ago.
The dates range widely: from roughly 300 CE to 1500. The pass saw peak use around 1000—the late Viking Age, when the population of Norway’s mountain valleys was growing and the seasonal movement of livestock to high-altitude pastures—transhumance—was part of the economic cycle. Lendbreen linked settlements, summer grazing grounds, and probably hunting areas. Calling it an “international trade corridor” would be a stretch: the archaeological material points to a regional route serving the local economy. People drove livestock, carried supplies, hauled cargo on sleds between valleys on either side of the ridge. After 1350—the time of the Black Death, which by various estimates wiped out up to two-thirds of Norway’s population—the traces on the pass nearly stop. The road fell empty, and the ice closed over it like a book no one had finished reading.
Among the thousands of objects collected at Lendbreen and across the wider region of Norwegian glaciers are finds that take archaeologists’ breath away. A woolen tunic around 1700 years old is one of the oldest surviving garments in Scandinavia. It was simply lying in the ice: someone lost it or discarded it on the pass in the 3rd or 4th century CE. The fabric retained the structure of its weave, making it possible to judge spinning and weaving technology long before the Viking Age. It is a window onto textile production that is usually sealed shut—fabric in the ground survives for only a few decades at best, unless peat or a desert climate preserves it.
Other finds take us further back still. Some objects from the wider region date to several thousand years BCE—the traces of hunters and herders who used mountain passes long before Scandinavians learned to forge iron. Arrows with stone and bone points, fragments of skis, scraps of rope—all of it survived the millennia only because it happened to be in the right place at the right time, when the ice grew and sealed the surface.
The count of finds at Lendbreen runs into the hundreds: around eight hundred objects, plus roughly one hundred and fifty bones and antlers and more than a hundred stone cairns. Across the country, glacier archaeology has already yielded thousands of artifacts—by 2024, 4500 finds from seventy sites had been documented in Innlandet alone—and the flow has not dried up: every warm summer exposes another strip. Sleds, skis, horseshoes, arrows—the inventory of mountain life, which left more traces in the ice than in written sources. The sagas describe kings and battles; the ice preserved dung and mittens.
The work of glacier archaeologists bears little resemblance to classical excavation. There are no test pits, no stratigraphic sections, no months of painstaking work with a brush. There is a melting ice edge retreating by meters each season, and a strip of exposed ground where artifacts lie on the surface or are barely covered by gravel. The task is to find and document an object before a stream carries it off, a rockfall crushes it, or the freeze-thaw cycle destroys it.
Lars Pilø and his colleagues use satellite imagery to monitor ice patches: it allows them to track which areas have retreated over the winter and where new finds are likely to appear. Reports from local residents—herders, hunters, tourists—serve as a second channel of information: someone on the ground notices what a satellite cannot make out. When a report comes in, the team heads out for an urgent field survey—sometimes that same day, sometimes a week later, but always knowing time is short. At nearly two kilometers up, weather changes in hours: morning fog gives way to sun that melts the ice at an alarming rate, and by evening snow may settle in. Archaeologists work in hiking boots, with GPS receivers and cameras, marking the coordinates of every object. Context—where an artifact lay, what was beside it, how deep it was below the ice surface—is no less important than the object itself. Without context, a woolen mitten is just a mitten. With context, it is a point on the map of a route.
I cannot shake the thought that this is archaeology inside out. Usually, an archaeologist comes to a site that is not going anywhere and methodically opens it up. Here, the site opens itself—and begins to fall apart while the archaeologist is still tying their laces. The program has been operating since 2006, received permanent funding in 2011, and in that time its methods have grown more precise: drones appeared for aerial survey, melt-prediction models improved, and a network of local informants was established. But the rate of melting is rising too, and the balance between what the ice reveals and what can be collected remains fragile.
The phrase “global warming is the best archaeologist” is pretty, but false. More accurately: global warming is a burglar who cracks a safe and scatters its contents in the street. Yes, we can see what was inside. But the contents are getting soaked in the rain, losing value with every hour.
Organic materials—wool, leather, wood, bone—enter an aggressive world once removed from the ice. Bacteria and fungi, inactive at subzero temperatures, begin breaking down proteins and cellulose. Ultraviolet radiation is more intense in the high mountains than in the valley, bleaching fabrics within weeks. Meltwater leaches water-soluble components—tannins from leather, resins from wood. Mechanical forces matter too: rocks shifted by melting ice grind fragile objects to pieces. To say that a find dies “in a matter of days” would be too categorical—the rate of decay depends on the material, the weather, and the specific location. But one field season is a realistic horizon for the loss of an undocumented artifact. A wooden arrow shaft that lay for a thousand years in stable cold can crack and crumble over the course of a summer once thawed.
The scale of the problem makes everything worse. Lendbreen is one pass. Norway has hundreds of ice patches and snowfields, each potentially containing artifacts. The Glacier Archaeology Program team physically cannot survey every site every season. Priorities have to be set: which glaciers are melting faster, where the likelihood of finds is higher, which routes were more active in the past. It is triage—a medical term for sorting casualties by urgency—applied to archaeology.
The most insidious loss is not the physical death of an artifact, but the loss of context. An object picked up by a случайный tourist and brought to a museum survives as an object, but loses its place in space. We no longer know whether it lay beside dung—meaning near the trail—or off to one side—meaning someone had left the route. We do not know how deep in the ice it was—and that indirectly indicates its age. We do not know whether other objects lay nearby, objects that together might have formed a story: a mitten beside a shoe and a broken sled fragment is an accident on the pass; a mitten alone is simply a loss.
What strikes me is how deeply archaeology depends on the relationships between things. A thousand objects laid out on museum shelves without coordinates are a collection. A thousand objects with precise coordinates, tied to the terrain, to stone cairns, to traces of shelters, are a map of human activity in the mountains over a millennium and a half. The difference is like that between a heap of letters and a text.
The program of Pilø, Finstad, and Barrett continues its work. Every summer the team returns to Lendbreen and neighboring sites, every summer the ice retreats farther, every summer a new layer of the past is exposed. Papers in academic journals appear regularly, finds are transferred to museums, data are entered into databases. But the pace of melting is set by the climate, not by expedition schedules. In warm years, the ice gives up more than the team can process. In cold ones, it pauses, and archaeologists have time to work through what has accumulated. Forecasts for Scandinavia’s ice patches promise no stabilization: most will continue shrinking over the coming decades. That means the stream of finds will grow, and then—when there is nothing left to melt—stop forever. Between those two points is a window into which all the work has to be squeezed. How wide it is, no one knows, including the researchers themselves.