Gingerly plucking burned pads of prickly pear cactus out of the hot coals and rocks in the pit oven they had dug open moments before.

"It smells like molasses," said Alston Thoms. A cowboy turned Texas A&M University archaeologist, he was commenting on the latest effort to reproduce a cuisine that dates back 6,000 years. That evidently was a time when early man in North America began to make a crucial, if little studied until now, change in lifestyle.

The pit oven was one of several dug on the edge of the presently receding Amistad Reservoir, which lies a bit east of where the Rio Grande and the Pecos River meet on the Texas-Mexico border.

It is an unpromising place, the Lower Pecos, with rolling desert-like terrain stretching endlessly in every direction and covered with thorny and spiky plants — agave, acacia, prickly pear. The soil, if the dry stuff can be called that, is littered with chunks and pebbles of white limestone, giving the landscape a bleached, dead look. It doesn't seem like it would be the number one choice for hunters and gatherers of any era. In fact, the ancient Indians tended to take up residence in the region's countless deep canyons where, among other activities, they spent a lot of time hauling limestone rocks into their caves and burning them.

All around the world, archaeologists have encountered burned rocks, rocks cracked by fire, pestiferous junk heaps of rocks in which they would dig around, pushing them aside, looking for artifacts — arrowheads, scrapers, what have you. All these darn rocks were in the way.

By a decade ago, however, some archaeologists had reinterpreted the rocks and now regard them as artifacts, part of what they call a "feature," a sign of a particular activity, in this case ovens that were usually in the form of pits. The pits were lined with rocks, over which a fire was made; more rocks were added, and then food, which could have been sealed in with dirt and baked, or wrapped in moist plants and steamed.

Around 7500 B.C., the landscape of the Lower Pecos had become desert-like. This meant that people were probably forced to change from a diet of mostly meat to one of mostly plants. But in a place like the south-central plains of Texas, the plants are downright hostile.

Lechuguilla, for example, which is a kind of agave (or century plant), contains an abundance of saponin, a steroid compound that is related to cortisone and is linked to a sugar. The sap from its lethal-looking pointy leaves can be used as arrow-tip poison or can be dumped into a creek to kill fish. But careful examination of some 300 human coprolites in one south Texas cave showed that, beginning about 6,000 years ago, lechuguilla was tops in popularity for a long time in the cuisine of these ancient Indians. Also high on the list of the some 50 plant species on the menu were sotol (another succulent), prickly pear cactus and wild onion.

For most of us, the picture that emerges from the existence of the pits and the analysis of the desiccated stools would be sufficient, but not for modern archaeologists. For them, from such lowly facts a thousand questions bloom.

What exactly were they cooking in those pits--lechuguilla? If so, why? And what else? It seems like an awful lot of work. How much work? How much rock and how much fuelwood does it take to produce and store how much heat to cook how much lechuguilla?

To answer such questions, you can look at the archaeological record and guess. You can also look at what tribal people today do in this regard--the ethnographic record. The Apaches, for example, still make pit ovens to cook agave. Another way, however, is to do it yourself, an approach known as middle-range research.

This is why the University of Texas and Texas A&M brought some 40 hot-rock archaeologists to a campground near the Amistad Reservoir last February. They did not make a pit oven the same way as the ancient Indians. The archaeologists dug the pits through a layer of caliche (crusted soil), and as one digger said, the ancient Indians "had limited access to pickaxes."

Besides that, the ancients didn't weigh anything that went in the pit, but here every rock was weighed, and all the firewood. So was the food and the cactus pads that were put on top of the fire and over the food to shield it. From the interstices of the rocks, the pads and the food, thin tubes of thermocouples snaked out to a meter attached to a computer in the front seat of a pickup. This was not, you might say, your mother's pit oven.

Several archaeologists had been preparing lechuguilla for the pit. This entailed chopping off the 30-odd fleshy leaves to get to the plant's heart much as a cook might prepare an artichoke. They hacked away strenuously with a variety of tools that were unavailable in the Holocene--machetes, even a saw--trying to achieve the neat 45-degree angle found in the archaeological record. While no one timed this effort, it took more than a few minutes and a lot of energy to lop off all of each plant's leaves. At this point, the grand guru of middle range theory, Lewis Binford of Southern Methodist University, demonstrated another way. With a rock as a hammer and a log as an anvil, he bashed each lechuguilla leaf, bruising it. Then, with a knife, he sliced them off quite easily.

Everyone agreed that this would seem to be the most efficient way to get to the lechuguilla's heart. Binford spent much of the weekend in conversations with individuals and small groups, always pressing, always challenging. "If we already knew exactly how to do this right," Binford said, "then we couldn't learn anything from it. You don't know what's germane until you spread yourself a little thin, till you put your knowledge in jeopardy."

The scientists concurred that burned-rock technology represented what they call "intensification" — a shift to a highly intensive means of using the landscape to survive. Driven perhaps by climatic shifts or population pressure, people were resorting to increasingly costly food-processing practices.

One scholar, LuAnn Wandsnider of the University of Nebraska, had recently published a seminal paper on the chemistry of such cooking practices, what could be thought of as "The Joy of Paleo-Cooking." Wandsnider concluded that our own practice of spending tremendous amounts of energy (often in the form of fossil fuel) to process a comparatively small quantity of food "is an ironic continuation of trends that began in the distant past."

Seeking to present these ovens in a slightly larger focus, Binford offered the scenario of a band of people sending out some young boys to alert relatives while at home the adults dig a big pit oven (some of the ovens that have been discovered measure as much as 12 feet across). For several days they have been out collecting lechuguilla and other plants. The extended family gathers from far and wide, the oven is filled, and they all talk about what's going on in the landscape — where the deer are, maybe, where the acacias are putting out seed.

So here, 6,000 years later, much the same thing is happening. In an age of e-mail and cellular phones, it seems that one of the most efficient ways to progress is still to call in the troops from here and there and chat it up while standing around a steaming, smoking pit oven.

On Sunday, two days after the meeting had begun, the time had come to put theory to the test. First to open a cooking pit was Phil Dering, an archaeobotanist with Texas A&M's Center for Environmental Archaeology. The lechuguilla had been baking for 36 hours, but the ideal cooking time, he confessed, would have been 48 hours, to break down all the saponins and free up those tasty sugars.

Once the pit was opened and the cactus pads plucked off, Phil offered around the much-awaited delicacy.

"It has a bit of a bitter taste still, Phil."

"It's a little soapy still, Phil." "This one is sweeter, like brown sugar maybe."

"You're not going to be taking a plane trip right away are you?" Phil asked one recipient who still was savoring the acrid, slightly mushy green stuff. "If it's not cooked enough," he explained, "it makes a great laxative."

"Hey, thanks a lot, Phil."

"Compliments to the chef."