By Lawrence Biemiller

Range Creek Canyon, Utah

"Just shimmy up here," Cody Chamberlain said, nodding at a two-foot-wide space between flat-sided rocks. The taller rock was maybe eight feet high, and Isaac Hart was already standing on top of it. He reached down to help me climb up.

When I stood and looked around, the view was magnificent. The high-desert canyon stretched as far as I could see. Below, grasses covered the canyon floor and trees marked the course of the creek, but at higher elevations vegetation was sparse. The canyon rim was still far above us, and in many places the canyon walls were sheer drops.

But the view wasn't what the two archaeology students had brought me up here to see. Slipping past rocks and windblown juniper, we came to what looked like the foundation of a small, round structure. Mr. Hart and Mr. Chamberlain had just finished the University of Utah's six-week archaeology field school here, and they were teaching me to use the correct terminology—"circular alignment of rocks"—instead of words like "house" that might prejudice future researchers. A few moments later we caught up with Nate Anderson, a teaching assistant at the field school this summer. He was checking up on a nearby rock alignment where shards of pottery peeked through the sandy soil.

The rock alignments we were looking at are among more than 400 sites in Range Creek Canyon linked to the Fremont people, who appear to have grown corn here from about AD 900 to 1200. Then they vanished, for reasons that archaeologists are still struggling to explain. Besides rock alignments, the Fremont left behind cutting tools, projectile points, pottery fragments, the remains of meals, petroglyphs (drawings scratched into rocks), and—especially mysterious—some of the planet's most inconvenient granaries, many of them built into all-but-inaccessible rock walls.

Duncan Metcalfe, the associate professor of anthropology who directs the field station here, had driven me in the day before. It's a five-hour trip from the university's campus, in Salt Lake City, so he had time to run down Range Creek Canyon's recent history.

Its remoteness has defined the canyon at least since the first white surveyor came through, in 1883 or 1884. What that surveyor found, Mr. Metcalfe said, was ideal grazing land, with good soil and a creek that runs year-round. But the canyon was, and is, hard to get to. Even today the last two and a half hours of the trip are spent on a rugged, one-lane dirt road that zigzags up steep mountainsides to reach a 9,000-foot-high pass, then drops down into the canyon and meanders from one side to the other, twisting around outcrops here and splashing across the creek there. In wet weather, Mr. Metcalfe said, the drive is harrowing.

Archaeologists identified the Fremont culture—named for the Utah river around which the first sites were found—in 1931. Around that time, an expedition organized by Harvard University's Peabody Museum of Archaeology and Ethnology located a handful of Fremont sites in Range Creek Canyon, but Mr. Metcalfe didn't see the canyon himself until almost 10 years ago. The Wilcox family, which had controlled the canyon since the 1950s, had just sold their ranch to the state. "That first spring, when I drove down with Waldo Wilcox, I wasn't going to get involved," Mr. Metcalfe recalled. Then, he says, "I spent 14 hours in the canyon and dropped everything."

On his next visit, he saw 77 archaeological sites in a week. "This family knew the canyon like the back of their hand. I bet 90 percent of the sites we've reported they knew about." There were 420 sites at last count, he said, adding: "We've barely scratched the surface." Last year the university and state officials signed a lease that gave the field station 1,520 acres spread along 12 miles of canyon.

"The thing that strikes archaeologists is how pristine it is," Mr. Metcalfe said. The Wilcox family never owned the whole canyon—much of it was public land on which they had grazing rights—but the key to ranching in the West, according to Mr. Metcalfe, "was to own as little as possible and control access to as much as possible." For a few years the Wilcoxes ran hunting tours in the canyon, but even then access to it was strictly limited. As a result, the archaeological sites went unstudied and undisturbed—the granaries were not kicked in, and the petroglyphs have no graffiti.

Now that the Wilcoxes are out of the picture, members of the public can apply for day passes to visit the canyon, but only on foot or horseback, so visitors are few. Fear of vandals, Mr. Metcalfe said, is one reason that he hasn't made an effort to improve the road coming in.

Growth and Collapse

As we worked our way down from the pass, he told me about the Fremont. They ranged across the Southwest and are distinguished from other groups by the trapezoidal anthropomorphic figures that appear in their rock art, by moccasins made from the skins of deer or sheep legs, and by having grown corn in addition to hunting and gathering.

"The typical house was a pit house with a wooden superstructure and wood and earth as a roof," he said, but the canyon has "a huge variety of structures." Radiocarbon dating on samples from sites excavated in the canyon so far has produced 24 dates, he said, of which 22 fall in the 900-to-1200 range. "One piece of basketry dates to 400, and one cache of wild tobacco that was under a rock ledge is about 1500." But he is careful to say that radiocarbon dating offers only estimates of age, and an old archaeology standby—counting tree rings—is of limited use here because juniper trees don't show much variation from one year to the next. So future research will include trying some additional dating techniques, such as subjecting soil samples to carbon dating and using mass spectrometry to study the tree rings.

The evidence available so far about Fremont habitation, Mr. Metcalfe said, could mean there was "a small population over 12 centuries or a large population over three." He leans toward the latter.

He believes people moved in when farming was very productive, and then the population grew and the climate changed. Bones found in cooking sites indicate that the Fremont at first butchered whole animals near where they were cooked, but later cooked meat that had been butchered elsewhere and brought to the sites—suggesting that they had to travel farther to find food. Still later, he said, there was "a switch from big game to smaller animals."

"I'm confident that the Fremont collapsed as a consequence of overpopulation and climate fluctuation," he said. The Southwest has "megadroughts" every 100 to 150 years, he said. "About 1300 they got a drought that really knocked them down." He thinks that "when things got bad," some of the Fremont may have moved up to high ground—among the 420 sites are some, along the tops of ridges, that are almost impossibly inaccessible.

Eventually we reached the old Wilcox ranch, a cluster of about a half-dozen buildings, including a house that the Wilcoxes painted pink. That's where the field-station manager, Corinne Springer, stays, as well as where meals are prepared. Ms. Springer, an archaeologist, is key to the field station, as adept at spotting Fremont sites as she is at regrading roads with the tractor or baking ever-popular lemon cookies.

Field-school students and teaching assistants, along with Mr. Metcalfe and his wife, Frances, stay in tents scattered across a pasture just below the pink house. The tents share the pasture with the field station's two horses, which are necessary because mechanized vehicles are not allowed off the road in much of the canyon. If field-school students work at remote sites, the horses carry in water and other supplies. The field station has a small solar array that powers Ms. Springer's computer, and a generator can be fired up to run laptops in the research lab or an old-fashioned washing machine with a wringer on top. There is no Internet connection or land line or cellphone service, although Ms. Springer has a satellite phone.

The field school operates only during the summer—usually for eight weeks, but this year just for six, because of budget cutbacks. The field station, however, is open from whenever in the spring Ms. Springer can get through the snow on the pass until whenever she reluctantly agrees to leave in the fall. In addition to field-school students, there are visits by other archaeologists, by researchers in other disciplines, and sometimes by groups of artists or others. Mr. Metcalfe is trying to broaden the field station's appeal.

"I've been an archaeologist for 30 years, and logistically this is the most challenging thing I've ever done," Mr. Metcalfe said the morning after we arrived. The field school is essential, he said, because "there's no way to teach field methods in the classroom—to learn how to excavate and survey requires doing it."

A History in Corn

But running the place is real work. Mr. Metcalfe and Ms. Springer were planning to spend the morning undoing some of the damage caused by a flash flood the afternoon before we arrived—a flood that had carried away the ranch's metal bridge over the creek, erased the road in several places, and buried one of the dams that held back water to irrigate the ranch pastures. "I've got to go do some triage," Mr. Metcalfe said as he sent me off with the three students. "The president of the university is coming through with a donor trip."

So it was Mr. Anderson, Mr. Chamberlain, and Mr. Hart who took me to see several Fremont granaries, adobe structures clinging to sheer rock walls or tucked into cracks in cliffs. One was perhaps 1,000 feet above the level of the creek. It was about six feet tall and six or seven feet wide, with a capacity of "a couple dozen bushels of corn," Mr. Anderson said. Two others were somewhat closer to the valley floor, but still very difficult to reach.

We talked about other foods the Fremont would have had access to—bighorn sheep, mule deer, rabbits, beans, squash, prickly-pear fruit, sego-lily bulbs, Indian rice-grass seeds, and pine nuts. Mr. Hart, who has planted two rare varieties of corn by the ranch buildings to see how they fare, spoke of foodstuffs in terms of how many calories a person would need per day. Pine nuts, he said, have "awesome calories," and the tiny rice-grass seeds would be easy to harvest at the right time of year by shaking the grass over a basket: "A talented collector in a good patch could collect 4,000 to 5,000 calories in an hour. And rice-grass seed is delicious roasted."

That afternoon Mr. Metcalfe talked about why the Fremont might have built their granaries in such remote places. A thunderstorm had brought everyone back to the ranch complex, wet and muddy—"Helluva way to run a desert," Mr. Met­calfe muttered—and they gathered in folding chairs under a tree outside the pink house. Mr. Metcalfe said that Shannon Arnold Boomgarden, the field station's assistant director, had combined high-resolution photos and topographic maps to create a computer program that "can fly you through the canyon."

"Her argument is that the granaries are all highly visible—they can all be seen for something like a mile," he said. So perhaps the idea was to protect the corn by putting it where anyone trying to steal it would have been in full view for however long it took to climb to the granary. Insurance companies today assume that burglars are unwilling to be in view for even as long as five minutes, so perhaps the same was true during the Fremont years in the canyon.

"Corn's a really tricky plant," Mr. Metcalfe said. "It's phenomenally sensitive to water stress during tasseling and silking. If the Fremont were facing climate variability, maybe they'd go up to the plateau above the canyon to hunt and gather." And if they did so, they may have depended on their neighbors to tell them if anyone had climbed up to raid their granaries while they were gone. "You start with the hypothesis that something makes rational sense"—something like building inaccessible granaries, Mr. Met­calfe said—"and then figure out under what circumstances it would be possible, and plausible, and sensible."

Still, archaeologists' work here has only just begun, and Mr. Met­calfe sees no point in rushing it along, in part because technology offers researchers more and more options every day.

"When we excavate a site we destroy it, even though we've recorded it," he said under the tree. "In the course of their careers, Cody and Isaac are likely to have techniques available that I've never dreamed of. So, all else being equal, we'd like to leave sites undisturbed. We only excavate to train students and to answer specific questions."

And the canyon has one kind of site he has stayed away from—burials. Even though it's not clear that the Fremont have any direct descendants among current American Indian tribes, the 1990 Native American Graves Protection and Repatriation Act involves more red tape than he's willing to deal with.

"What we don't know about the Fremont isn't a consequence of not having excavated enough sites," he continued. "What Isaac and Cody will discover will be a consequence of new techniques or new theories. If all I do before I retire is preserve this site for the future, I'll retire a happy man."

After dinner another thunderstorm approached, and I climbed a nearby knoll with Mr. Anderson, Mr. Chamberlain, and Mr. Hart to watch a spectacular lightning display. Mr. Chamberlain saw lightning set a tree on fire way down the canyon, but then it began to rain, and it rained for much of the night. In the morning Ms. Springer said another flash flood had wiped out all of the previous day's work. "Sometimes," she said, "a girl just wants to be alone with her shovel and a whole lot of F-words."

The university's lease on Range Creek Canyon runs for 30 years. That's a point about which Mr. Metcalfe is especially pleased, because nowadays archaeologists often find themselves working on tight deadlines as construction crews loiter nearby, waiting to start office buildings or housing developments. On the other hand, as Mr. Anderson told me while he and I dragged part of the bridge back across the pasture, "Range Creek may not be quick to give up its secrets."