Exploring America’s Earliest Rock Art



Exploring America’s Earliest RockArt 


JackSteinbringUniversity of Wisconsin-Oshkosh and Ripon College


The recent discovery that the Winnemucca Petroglyph Site (Fig. 1)

Fig 1




in Nevada may date to as early as 14,800 years ago has prompted a review of other instances of early rock art in North America. The Winnemucca Site contains deeply eroded petroglyphs including panels of what appear to be randomly produced cupules.

Cupules, in general, have been assigned an early context throughout the world, including North America where Parkman (2007: 1) has viewed them in early contexts, as well as reminding us that cupules have been produced in modem times in California where they are part of fertility rituals. Thus, it becomes critical that physical evidences of antiquity be established.

In various places these arise from water level relationships , degree of patination and repatination (Heizer and Baumhoff 1962:l 9) and sometimes on the basis of the iconography itself. From time to time the relationships between certain forms and how they are placed on the rock formation provide clues. From these instances we can see which forms occupy the most prominent and “best” surfaces, and which forms came later and were forced onto less desirable locations. In Canada, for example, the Lake of the Woods Style (Fig. 2) precedes others at the Mud Portage Site, while the much more recent Clearwater Bay Style (Fig. 3) is found on the peripheries (Steinbring, et al: 1987:3).

Steinbringfig3The Winnemucca Site is interesting because it is not far from another Nevada site which has attracted the interest of rock art historians. This is the Grimes Point Site (Fig. 4) which contains random cupules that have undergone extreme erosion (Heizer and Baumhoff 1962: plate 1). Steinbringfig4Another purported early site is the Grapevine Canyon Site in Nevada (Fig. 4a), once covered by sediments known to be at least 10,000 years old (Heizer and Baumhoff 1962:28).












Looking more widely, a site in Idaho also offers random cupules, again very deeply eroded. This is the Wees Bar Site (Fig. 5) on the Snake River (Tobias 1981).



Actually, in this case, amateur excavations revealed cupules and lines beneath the soil mantle, one of the very few North American examples of such a situation. In all known cases, however, the engravings are almost certain to have been prehistoric.
Another area in which cupules figure in very early rock art is in the Northern Plains in the region encompassing southeastern Alberta, southwestern Saskatchewan and northeastern Montana. Here a phenomenon known as “ribstones (Fig. 6) is found.

Fig 6

These consist of large glacial erratics with cupules engraved on them as well as lines which are executed so as to represent the ribs of large animals, mostly perceived as bison. In one instance, the engraved animal is thought to represent a musk ox (Parkman 2007: 9). Investigators familiar with the ribstones have judged them to be ancient (Parkman 2007: 10). The ribstones fall into the Pit and Groove Style (Heizer and Baumhoff 1962:19) which has long been regarded as the earliest identifiable rock art style in North America (Parkman 2007:1). Tentative evidence of this has emerged from the excavation of a ribstone site near the village of Scapa, Alberta (Head 1988). A pipeline project there prompted the government of Alberta to require the pipeline company to undertake test excavations for the determination of threat to the site since the pipeline, as planned, would pass directly through it. The resulting excavations yielded Paleo-Indian projectile points beneath a level of volcanic ash from the Mazana eruption of 6,700 B.C.E.

Unfortunately, the excavations were discontinued at this point, and the monolith itself (Fig. 7) was stolen!

Fig 7
The ribstones precisely occupy an area (Fig. 8) critical to the peopling of the North American Mid-Continent. This is the exit of the Alberta Corridor, one of the main avenues of entry.

Fig 8

Another exists along the Northwest Coast, where again, cupules constitute the most widespread marking phenomenon of prehistoric rock art there (Fig. 9).

Fig 9
There is no question of the early occurrence of cupules along the northwest coast, since many of them are found on large boulders which are covered during mid-tide. This is in accord with the fact that, during glaciation, water levels were much lower than they are now – at times over 400 feet lower. There is developing, at this time, assertions that the coastal route was initially used by the first Americans in their movement into the New World. This approach would have required watercraft because the Northwest Coast is deeply fjorded. Excavations at the Paisley Caves in South Central Oregon yielded fossilized human coprolites dating to 14,000 years B.C.E. This suggests that passage from the Northwest Coast was quite early. Since watercraft had been invented and used as early as 40,000 years ago elsewhere, the assertion of Northwest Coast entry is not unreasonable.

However, the Alberta Corridor as a natural passage from Northeast Asia is known to have existed for millions of years (Cushman 1966:3). Paleontologists have spent many decades researching the movements of early animal life along this corridor, and it obviously posed the central route of mastodons, mammoths, bison, extinct forms of the horse, camel, etc. – all animals critical to land dwelling hunters and foragers proceeding in search of them from Northeast Asia. During the glacial maxima, the connection between continents was as much as 1,200 miles wide. Those who have argued more ardently for an exclusive coastal entry have failed to consider the priority posed by North American paleontology. Needless to say, a coastal route would not have been congenial for elephants and bison, grassland animals of the steppes and plains.

Virtually all DNA studies focused on the entry of human groups into the continental interior show no coastal passage, and there is some evidence beyond the Pit and Groove rock art style that support the interior corridor theory. This exists in the form of actual human remains. In Minnesota, virtually in the center of the North American continent, no less than three sites have yielded human remains in very early contexts. These are the Gold Island skulls (Steinbring 1974:67), the Brown’s Valley Burial (Jenks 1937, Wormington 1957:143), and the Pelican Rapids skeleton (Jenks 1936, Wormington 1957:232). The Gold Island skulls were discovered in the Boundary Waters Canoe area bordering Ontario. They were accidentally found by “rained out” fishermen sheltering beneath their overturned canoe. Idling their time away, one started scratching in the dirt with his camp spoon. This effort disclosed a human skull from an intentional burial dug into a terminal glacial feature. The characteristics of this skull are of some interest. The physical anthropologist who first examined it, and two other skulls found shortly later, described the length-breadth ratio as the most extreme dolicocephalic index on record for North America. There was also thickness of the cranial bones. These features were also expressed by the Pelican Rapids female (first called “Minnesota Man”). She seems to have drowned in a proglacial water body (Wormington 1957:236). Her bones had become crystalized, and could not be used for radiocarbon dating. There is no question of her very substantial antiquity.

The Brown ‘s Valley skeleton was also found under conditions of substantial antiquity. In this case, the skeletal material was part of an intentional buriFig 9bal into a newly formed gravel ridge associated with a lake formation provisionally dated to 18,000 B.P. (Wormington 1957:144). The artifacts recovered included a beautifully executed projectile point with parallel oblique flaking. It is now called the Brown’s Valley type (Fig. 9b). The skeletal material has been recently re-dated and a date of over 10,000 years B.C.E. has been obtained.

Fig 10
Another early burial of note was excavated at the Bjorklund Site (Fig. 10) on the west bank of the Whitemouth River at its confluence with the Winnipeg River in Eastern Manitoba. In this case the burial was a female primary burial , flexed on the right side and placed in a shallow oval pit. She was covered with red ochre and had a broken chert biface thrown in as a funereal accompaniment , along with crossed bison rib bones and a large asymmetric biface over the chest. The cranium was removed and “excavated ” in the lab, and in the course of this effort two columella beads were recovered from the neck area. They had been held there by her hands. All the others were missing. Many of these features (almost all) characterized another early burial in Northern Colorado. This one, called the Gordon Creek Burial, was a primary burial of a female placed in a shallow pit grave. She had elk teeth beads around her neck, most of them also missing. Broken bifaces were present as well, and the body was covered with red ochre. The Gordon Creek Burial was radiocarbon dated to 9,750 years B. P. (uncorrected) (Breternitz, et al, 197:172). There can be no doubt that the two burials reflect a precisely identical funereal pattern and probably occupy a similar time zone. Unfortunately , the Manitoba burial could not be dated accurately because a large rodent tunnel had penetrated the skeleton, carrying contaminants into and around the body cavity.

It is interesting to observe these early northern burials, most of them near the terminus of glacial conditions. The pattern of initial human movement from the exit of the Alberta Corridor may illuminate this distribution. It is definitely in accord with some of the DNA evidence of human trajectory in very early times, and it may be seen to relate to the ribstones. They may reflect the artistic dimension of interior America ‘s earliest inhabitants. As these groups spread out from the confines of the ice lined corridor, some proceeded south and turned west venturing into the few passages allowing penetration of the mountain barrier, eventually reaching Winnemucca , Wees Bar, and Grimes Point.

Some groups, from DNA evidence, moved east along the glacial front. This idea is sustained by the finding that the Ojibwa Indians always fall into the groups thought to be descendants of the first Americans. Virtually all Algonkian speakers are seen to be in this category and the most remote ones, like the Nascapi (Speck 1935:78) are probably the closest genetically. This is because they have adjusted to the Pre-Cambrian Shield, the last great refugium of North America. Still without roads in most of it, it remains an ideal place to escape extensive interaction. The idea tested here is that the first Americans, with a basic hunting and foraging economy, followed the animal movements through the Alberta Corridor, and skirted the still existing glaciers to the east.

In one case these early groups occupied a huge peri-glacial void in northeastern Minnesota at about 14,000 years ago (Steinbring 2004:140). From here, they moved into a further peri-glacial embayment in northwestern Ontario (Fig. 11).

Fig 11

And at this time, they became trapped by renewed glaciation. Their sites were found on the southern flanks of moraine deposits overlooking Glacial Lake Aggassiz from the north in Ontario. As the lake drained, they moved south and onto Lake-of-the-Woods where they occupied the Mud Portage Petroglyph Site. They then interacted with other groups already established to the south.

Somewhere between southeastern Alberta and Lake-of-the-Woods, Ontario, and over a span of at least 3,000 years, the style of art had changed completely, from non­ representational abstract art in the form of cupules and lines, to naturalistic representational, contiguously pecked figures. This mode of portrayal is seen all across the subarctic, from the Early Hunter Style of Wyoming to the Lake-of-the-Woods Style in Ontario. This is in accord with the changing settlement pattern from non-residential wandering, to a central based wandering pattern with more intensive regional exploitation. The elephants, for example are gone (for whatever reason) and smaller animals are in, including the bison with known migratory limits. The transition from Paleo-Indian to the Archaic Cultural Tradition is fairly smooth and long term. It is a continuation of central-based wandering with increasing efficiency to the point of enough leisure for the arts to fluoresce. Rock art remains representational, but there is much more of it. In some areas the style evolves more rapidly than in others, and gives the impression of fundamental change. But in the main it is the exquisite sculptures (Knoblock 1939) in the variety of the atl-atl weights, boatstones, and axes which mark the ultimate florescence of Archaic art and sees its continental spread.

That spread is directed to the south in Paleo-Indian times to even fairly deep interior continental locations . In central Wisconsin , the Hensler Petroglyph Site exhibits Paleo-Indian projectile point imagery, with the peculiar enhancement of twinning. This mode of perception is manifest in the defining attributes of Algonkian Culture wherein transformation characterizes mystical thought. Twinning imagery is a common feature of Algonkian decoration. Moreover, the Hensler Site, with its atl-atl image, and contiguously pecked naturalistic animals, is now known to be a t least 10,000 years old, and with the clear prospect of being potentially much older. Recent revisions of the timing of glacial recession there show the area to have been ice free by 15,000 years ago (Syverson 2011 :537). Since the depositions of Aeolian silt (löess) date to immediately post-recession times, and that this deposition lies above human marking on the rock formation , the prospect of very early occupation of the site is possible . Indeed, this timing suggests the chance that the earliest marking at Hensler may be pre-Clovis . This would not, however, be the first instance of pre-Clovis marking. The Gault Site in Texas offers the first known human marking during Clovis times. This consists of mobiliary marking at Clovis levels in the excavations there (Collins 2002:33). The iconography is fundamentally different from that posed by the early cupules and lines of the north and the representational forms aligned with late Paleo-Indian and Archaic of the north and Midwest. The Gault forms on small stone tablets are geometric, consisting mainly of parallel or dendritic lines.


A Clovis burial of a child dated to 12,600 years B.C.E. near the Yellowstone River puts Clovis within access of the area of the ribstones, and within drainages leading west past the mountain barrier. No rock art of an iconography attributable to Clovis times is known for the immediate vicinity.

The presence of deeply eroded cupule sites is known in deep interior areas to the east however. Such sites are known in Missouri (Diaz-Granados and James Duncan 2000:63), as well as sites as far south as Georgia.

All this suggests that powerful cultural impulses propelled the early marking into these regions, without significant change, to what were then the remote peripheries of Pre-Clovis and Clovis life. In fact, the presence of Clovis, seen here as a descendent of
earlier groups, became well defined culturally and spread throughout the entire continent. In fact there are more Clovis remains east of the Mississippi than west of it (Mason1962:227)).

It is not a wild assumption that Clovis had a rupestral art (bedrock) component, and probably an easily recognizable one. Was the popularity of random cupules still significant, or were the Clovis people, themselves, already switching to the carefully executed animals of the early Archaic? It is high time science fixes the cultural parameters of these forms so that a continuum of aesthetic perception can be developed.



Author’s Note: A version of this paper was delivered to The Eastern States Rock Art Research Association at Natural Bridge State Park, Kentucky on April 5, 2014. It contains full documentation. Questions are encouraged.


References Cited


Breternitz , David A., Alan C. Wedlund , and Duane C. Anderson
1974 “An Early Burial from Gordon Creek, Colorado”, American Antiquity , Vol. 36, No. 1, pp. 170-182, Washington , D.C.

Collins, Michael B.
2002 “The Gault Site, Texas and Clovis Research “, The Athena Review, Vol. 3, No. 2, pp. 31-102

Cushman, Dan
1966 The Great North Trail: America’s Route of the Ages, McGraw-Hill , New York

Diaz-Granados, Carol , and James R. Duncan
2000 The Petro glyphs and Pictographs of Missouri , The University of Alabama Press, Tuscaloosa

Head, Thomas
1989 Final Repo1i: ASA Pro ject No. 88-163, Bison Historical Services Ltd, Calgary, Alberta

Heizer, Robert F. and Martin A. Baumhoff
1962 Prehistoric Rock Art of Nevada and Eastern California, University of California Press, Los Angeles

Jenks, Albert E.
1936 Pleistocene Man in Minnesota: A Fossil Homo Sapiens, University of Minnesota Press, Minneapolis

Jenks, Albert E.
1987 “Minnesota ‘s Brown’s Valley Man and Associated Burial Artifacts,” Memoirs of The American Anthropological Association , No. 49, Menasha, Wisconsin

Knoblock , Byron W.
1939 Bannerstones of the North American Indian, Published by the author, La Grange, Illinois

Loring, Malcolm J. and Louise Loring
1983 Pictographs and Petro gl yphs of the Oregon Country Part II, Southern Oregon, Institute of Archaeology , University of California, Los Angeles

Mason, Ronald I.
1962 “The Paleo-Indian Tradition in Eastern North America”, Current Anthropology , Vol. 3, pp. 227-283 .
Parkman, E. Breck
2007 Pit and Groove Antiquity in New World Prehistory Science Notes No. 89, California State Parks, Petaluma, CA

Prest, V.K.
1968 “Quaternary Geology” in Geology and Economic Minerals of Canada Dept. of Energy Mines and Resources, pp. 675-7, Ottawa

Speck, Frank G.
1935 Naskapi: Savage Hunters of the Labrador Peninsula, University of Oklahoma Press, New Edition 1977

Steinbring, Jack
1974 “The Preceramic Archaeology of Northern Minnesota”, in Johnson (Edit) Aspects of Upper Great Lakes Anthropology : Papers in Honor of Lloyd A. Wilford, Minnesota Historical Society, St. Paul

Steinbring, Jack, Eve Danziger and Richard Callaghan
1987 “Middle Archaic Petroglyphs in Northern North America,” Rock Art Research, Melbourne, Vol. 4 (No’s 1 and 2):3- 16, 150-158.

Steinbring, Jack
2004 “Elemental Forms of Rock Art and the Peopling of the Americas”, in Diaz and Duncan (Edits) The Rock Art of Eastern North America, University of Alabama Press, Tuscaloosa

Syverson, K.M., P.M. Colgan
2011 “The Quaternary of Wisconsin: An Updated Review of Stratigraphy, Glacial History, and Landforms”, in Ehlers, Gibbard and Hughes (Edits) Quaternary Glaciations – Extent and Chronology, A Closer Look, Developments in Quaternary Science, Vol. 15 pp. 537-552, Amsterdam, The Netherlands

Tobias, Nellie
1981 The Wees Bar Petroglyph Field, Southwestern Idaho, Boise State University Press, Boise

Wormington, H.M.
1957 Ancient Man in North America, Denver Museum of Natural History, Popular Series No. 4, Denver

Leave a Reply