Origins of Agriculture | Archaeological Research in Oaxaca, Mexico

Agriculture in ancient Mexico began during the Archaic period (10,000-2000 BC), when the inhabitants of the Valley of Oaxaca were still nomadic hunters and gatherers. These people lived in open-air camps, caves, and rockshelters.

In December 1964, Flannery visited 30 rockshelters near Mitla in the eastern Valley of Oaxaca. In late 1965 the Smithsonian Research Foundation awarded him a pilot grant that would allow him to test some of these sites (January-August, 1966). Flannery assembled a team that included geomorphologists Michael and Anne Kirkby, palynologist James Schoenwetter, botanist C. Earle Smith, Jr., fellow archaeologist Frank Hole (then of Rice University), and several graduate students. The Frissell Museum of Zapotec Art in Mitla rented the project a laboratory. Legendary Mexican archaeologist Ignacio Bernal provided Flannery with a list of 274 archaeological sites he had already located, and director José Luis Lorenzo personally undertook to see that Flannery’s project got a permit from the Mexican Institute of Anthropology and History (INAH).

During 1966 the project was able to excavate Guilá Naquitz Cave, Cueva Blanca, and the Martínez Rockshelter; an open-air Archaic site, Gheo-Shih, was added in 1967. Also briefly tested were Cueva Redonda, Silvia’s Cave, Cueva de los Afligidos and Edwards’ Painted Cave.

The most productive Archaic sites lay on lands of the now-abandoned Hacienda “El Fuerte,” about 4 km northwest of Mitla. This region of volcanic tuff cliffs and canyons produced evidence of a Paleoindian period and a sequence of four Archaic phases—Naquitz, Jícaras, Blanca, and Martínez.

The available palynological and zooarchaeological data suggest that the Late Pleistocene environment of the Valley of Oaxaca was cooler and drier than today’s. The mountains near Mitla evidently supported the pinyon pine, a northern Mexican species that can no longer be found in Oaxaca. The fauna from Zone F — the basal layer of Cueva Blanca — included the Texas gopher tortoise, another northern Mexican species that no longer lives in Oaxaca. Two Paleoindian fluted points, found within walking distance of Cueva Blanca, have been provisionally assigned to an El Fuerte phase for which there are still no radiocarbon dates.

Three radiocarbon dates from Feature 15 of Cueva Blanca suggest that the Paleoindian-Archaic transition took place between 11,810 and 9870 BC (calibrated). The oldest Naquitz phase level, Zone E of Cueva Blanca, dated between (cal.) 10,718 and 8304 BC.

Guilá Naquitz Cave produced 35 more dates for the Naquitz phase, suggesting that it lasted until (cal.) 6000 BC. The extraordinary preservation of plants in Guilá Naquitz —a very dry cave— was the project’s major source of information on the origins of agriculture. Bottlegourds and Cucurbita pepo squash were domestic by (cal.) 8000 BC. The occupants of Guilá Naquitz were also collecting runner beans (Phaseolus sp.) at this time, but those beans cannot be identified to species. In addition, evidence shows that pinyon nuts were still available near the cave, although they gradually disappeared during the Naquitz phase.

The available evidence suggests that during the Naquitz phase, foragers in the Mitla region underwent cycles of aggregation and dispersal. When resources permitted, groups of 25-50 persons camped in the open at Gheo-Shih, a 1.5 hectare site near the Río Mitla. At lean times of the year, these groups broke up into individual nuclear families of 4-6 persons who camped at rockshelters such as Guilá Naquitz. There are suggestions that most rituals were carried out when 25-50 persons were encamped together. For example, Gheo-Shih produced a ritual feature consisting of a cleared area between two lines of boulders.

Figure 2. A ritual feature in the basal level of Gheo-Shih.

Upper levels at Gheo-Shih belonged to the Jícaras phase, (cal.) 6000-4000 BC. The dominant projectile point of this phase was the Pedernales type, characterized by a concave-based stem. At roughly (cal.) 4355-4065 BC, Jícaras phase foragers left five primitive maize cobs at Guilá Naquitz Cave. These small cobs —with only two rows of kernels —- represent the oldest corn directly dated by AMS techniques.

Figure 3. Five early, two-rowed maize cobs from the Jícaras phase, Guilá Naquitz Cave. (Scale at right in millimeters).

All Mitla caves and rockshelters were excavated using a grid of 1 x 1 m squares; all artifacts were counted by squares, and all projectile points and other bifaces were piece-plotted. These practices made possible a whole series of spatial analyses by collaborators Charles Spencer, Robert Whallon, and Robert Reynolds. One could detect tool kits, “drop zones,” “toss zones,” and men’s and women’s work areas on many of the Naquitz phase living floors. Often the women’s work area included a hearth.

Blanca phase levels at Cueva Blanca, (cal.) 4000-2000 BC, revealed a potentially significant social change. While stratigraphic Zones E and D at Cueva Blanca showed the usual pattern of men’s work areas vs. women’s work areas and hearths, Zone C was different. This living floor had no hearth or likely women’s work area; it appeared to be a short-term men’s deer hunting camp. It is therefore possible that by the Blanca phase, the growing of maize, squash, and gourds had become so successful that the occupants of the Mitla region did not always disperse into small family groups during the leanest seasons. Now they were staying longer in their large open-air camps, from which small groups of men went out on brief deer hunting trips. The late Lewis R. Binford referred to this shift as one from “foraging” to “collecting.”

The Martínez phase represented the final moment of Archaic life in Oaxaca. During this brief period (cal. 2000-1800 BC), ground stone tools increased in number and variety while projectile points gradually disappeared. The Formative village societies that followed the Archaic virtually lacked projectile points.