By Jack Steinbring
University of Wisconsin-Oshkosh and Ripon College
The fact that rock art forms a tangible, physical phenomenon, entirely amenable to empirical analysis, has not been fully explored in the Americas. And, this is especially true for archaeological science. As an artifact, an element of rock art may at times be investigated to determine its cultural context and timing quite the same as in any other archaeological investigation. Not much has been done along these lines. Mostly this kind of archaeological approach has been pursued in the Old World, and then primarily in Paleolithic caves. Where rock art occurs in New World caves, a few efforts have been undertaken to connect the images with specific cultural deposits. Notable among these was Pictograph Cave in Montana (Mulloy 1958), the Hueco Tanks Site in Texas (Jackson 1938, Kirklund 1939), and several sites in South America (Guidon 1986).
All of these yielded useful information, and speculations about relationship were advanced. It is only now becoming clear, however, that specialized attention must be directed to these kinds of studies to include sites in the countryside where the rock art is fully exposed and there is no orderly cultural buildup as there is in caves and rock shelters. Perhaps one of the more substantial long-term clinical studies in the western hemisphere is that of the Gottschall Rock Shelter in Iowa County, Wisconsin. This site was continuously excavated, with meticulous precision, by Dr. Robert Salzer of Beloit College for more than 15 years. Dr. Salzer attempted to relate about 40 paintings on the rock shelter walls with specific prehistoric horizons revealed in the stratigraphy below them. He was able to show that most, if not all, of the paintings were produced by groups related to the Mississippian Culture and to that of the Effigy Mound builders of Wisconsin (Salzer 1987:466). The dates established place the rock art in the A.D. 600 to 1200 time period. The groups were probably horticulturists and they participated in many identifiable rituals including (besides rock painting) the construction of thousands of earthen mounds in the shape of animals and birds. There is evidence at Gottschall that these people brought ceremonial soils from elsewhere and placed it upon the rock shelter floors.
The most promising conditions for the archaeological investigation of rock art are when the rock art is, in fact, in direct contact with the cultural deposits. Such a situation is rare. In these cases the rock art on walls is usually covered by deposits, and we can only say that the carving or painting was done before it was covered up. So, we may have a period of thousands of years before deposits reached a level high enough to cover the rock art. Even when we find pieces of exfoliated fragments of paintings in the stratified deposits, we have a similar problem. We have not determined when the painting or carving was made, only the time at which it fell from the wall or ceiling and got covered up by the accumulations of everyday life. Dr. Salzer, in his careful scrutiny of the 2.0 cm. levels of his cave excavations found that pigment stains matching the paint on the walls had dropped onto the living floor at a certain level. This showed the exact stratigraphic/cultural point at which the pigment was applied! This, however, concerns only the drop or drops of pigment involved. It still does not tell us all we would like to know. It tells us only that a sloppy painter spilled some paint at one infinitesimally small point in time and culture history. Nonetheless, it is a whole lot better than most archaeologists have been able to do! And, in the light of science, we proceed from the known.
Outdoor sites are even more difficult. At the Mud Portage Site in Northwestern Ontario in Canada, stratified cultural horizons lie above a horizontal panel of petroglyphs pecked into a large basalt panel. Excavation of these deposits led to the clear understanding that the earliest deposits were newer than the dozens of delicately pecked animals and other figures on the basalt panel (Steinbring, Danziger and Callaghan 1987). As is too often the case, the lowest deposits contained virtually no "diagnostic" artifacts! However, the deposits were arrayed in a regular sequence, and a good estimate of timing halfway down the deposits was possible. This estimate was achieved, through the use of a number of archaeological and geologic criteria, at 5,000 B.P. An inference, based on some taxonomic data and the fact that the petroglyphs had been totally repatinated before deposits were let down on them, made a working hypothesis of 9-10,000 B.P possible. The research pertaining to a precise assessment of the archaeological context of this intriguing rock art is not finished. It might include the AMS dating of carbon particles on stone tools. Possibly, they were used in making the petroglyphs themselves. The tools were lying directly on the engraved surface, theoretically articulate with the carvings.
In southwestern Saskatchewan, Canada, several seasons of direct archaeological investigations of carved and painted monoliths were undertaken in the early 1990's. This work simply entailed the formal excavation of the base of the monoliths (Steinbring and Buchner 1997). The results are not much different from similar excavations next to decorated cave walls. The cultural deposits accumulate alongside the carvings, and in one case excavation revealed that the deposits covered several rock paintings not previously known! Just the same, very little cultural information was developed. For one monolith, at the Swift Current Creek Site, a horizontal panel of images consistent with that of known Archaic (5,000 B.P.) sites elsewhere was above ground. The excavations showed that some carvings were covered by the upper soil horizon, and that paintings were covered by lower ones. Interestingly a thin lens of pigment and tiny pecking tools was discovered on a plane which bisected a painting. This, of course, meant that the deposit came well after the execution of the painting. A tooth found in the articulated lens yielded a date of A.D. 6-800, consistent with a bison-hunting culture known as Avonlea. Still, does this prove that the painting was done by people from that culture? No. It does not, because the deposits containing evidence of this culture were laid down after the painting was applied. It is probably a good bet that Avonlea people made it, but there is absolutely no direct physical proof. And, what about the seemingly early petroglyphs on top of the monolith? Typologically they would be more than 3,000 years earlier than Avonlea. Nothing was found in these excavations that would support such an early date. And, many rock art authorities at the present time reject dates achieved by typological data alone!
In another Saskatchewan site, the Herschel Site, excavation around a richly carved monolith high on a hill, yielded a clear case of ritual offerings by a group whose deposits covered some of the carvings. The dates obtained from bone here yielded evidence of the Avonlea culture once again. Were they the perpetrators of the art? Who knows? All we can say is that they did deem the site to be sacred, and did perform rituals there, but, they may have started doing this thousands of years after the first of the carvings had been made! It is a well known fact that in the American Mid-Continent, historic Indian tribes considered ancient burial mounds to be sacred, and used the tops of them to bury their own dead at least 2,000 years after the first burials were made and the mounds were constructed over them. At Herschel, offerings continued well into the Historic Period. These offerings included a decorated tin lid, and small buckles like those commonly found on overshoes!
Thus, what we see in the application of direct archaeological studies of rock art sites is a whole range of complications which demand that persevering techniques be initiated. The best advise is to employ as many independent procedures of study as possible, and to deny none. This means that we do not throw out taxonomy; only that we use it in association with as many other approaches as will apply: stratigraphy, radiometric dating, geological and chemical studies, landscape factors, weathering, and anything else one can think of. Establish no priorities. Use everything!
1986 "A Seqüencia Cultural da Área de Sao Raimundo, Nonato,
Piauf, CL10." Revista do Mestrado em Historia. Universidade
Federal de Pernambuco, Recife: pp. 137-164
1938 Picture Writing Of The Texas Indians, University of Texas,
1939 "Indian Pictures In The Dry Shelters Of Val Veroe County,
Texas," Texas Archaeological and Paleontological Society,
1958 A Preliminary Historical Outline For The Northwestern
Plains, University of Wyoming Publications, Vol. 22, No. 1,
1987 "Preliminary Report On The Gottschall Site (47 IA 80),"
Wisconsin Archeologist, vol. 68, no. 4, pp. 419-472,
Steinbring, Jack, Eve Danziger and Richard Callaghan
1987 "Middle Archaic Petroglyphs In Northern North America,"
Rock Art Research, vol. 4, no. 1, pp. 3-16, vol. 4, no. 2,
pp. 150-161, Melbourne
Steinbring, Jack and Anthony P. Buchner
1997 "Cathedrals Of Prehistory: Rock Art Sites Of The Northern
Plains," American Indian Rock Art, vol. 23, pp. 73-84,