|
|
|
|
Current Research on Wisconsin Petroforms
by Jack Steinbring
The term “petroform” was first used by Dr. Peter Douglas Elias in the late 1960’s to describe the huge Tie Creek Site in the Whiteshell Provincial Park, Eastern Manitoba, Canada (Steinbring 1970). This site covers nine acres of typical Precambrian Shield country, an irregular expanse of the world’s oldest granite formations amidst the glacially produced rivers and lakes of central Canada. Seven boulder features, mainly geometric, were laid out on the open rock formations, at least 2,000 years ago, some of them perhaps 1,000 years earlier. All sizes of boulders, some weighing up to a half ton were arranged on open rock and linked together by boulder lines. The rediscovery of this site inspired regional explorations and, within a decade, 50 sites (all smaller) were recorded in eastern Manitoba (Danziger & Callaghan 1986:57). These ranged in shape from serpents and turtles to circles and rectangles, usually to be found near rapids on the regional rivers. Early explorers and traders had noted their presence in the 18th century, but no connection to any specific tribal group has ever been fully established.
Modest variations of what we now call petroforms have been known in the American Southwest since the 1930’s, the more spectacular ones created by the removal of desert pavements to form huge landscape productions of serpents, humans, and other forms. These have been called “geoglyphs,” a term which also embraces the occasional boulder arrangement. In the American Plains, boulder arrangements in the form of circles and lines, including those called “Medicine Wheels,” which have clear astronomical properties (solsticial and equinoctial alignments) were featured in early Plains rituals. Alfred Miller a 19th Century artist depicted one of these consisting entirely of bison skulls. Further to the east, as far as Iowa the inveterate recorder, T.H. Lewis (1889, 1890, 1891) found animal figures, humans, and circles made up of boulders laid out on the ground. Lewis was the first to “excavate” a petroform. In a regular circle, he noted the absence of some boulders. He probed these spots and found that the “missing” boulders were still there, buried by time.
There has been an upsurge in petroform research in recent years, some of it in eastern Wisconsin where J. Behm and H. Bender and their colleagues have recorded a number of sites in the vicinities of the Horicon Marsh and Kettle Moraine (Behm 1991:12, 1991:31, Behm et al 1989:28). Astronomical attributes are asserted for some of these, and their locations sometimes overlap with effigy mound sites. When not associated with effigy mounds (geographically), the loci of petroforms almost exclusively contain Archaic sites. From this, there is some reason to speculate that effigy mounds may be descendant from the petroform phenomenon. In re-examining old reports on effigy mound excavations, cases of pre-existing boulder features were found on the cleared floors. This might suggest that petroforms predated mounds at times.
One thing that has not been characteristically associated with petroform research has been the clinical archaeological dimension – basically, excavation. One relatively recent effort of this kind was conducted in Western Saskatchewan in Canada in the early 1990’s. There, excavations at the Swift Current Creek Site and the Herschel Site yielded vital information about the culture and context of both carved monoliths and a “boulder pavement.” The latter case involved the actual discovery by excavation of a circular pavement of uniformly sized boulders laid solidly together at around A.D. 600 in front of a richly decorated monolith. Next to it were many offerings, left up to as recently as the late 19th Century! The same site contains a long line of boulders stretching down a slope containing three carved monoliths, to Eagle Creek a quarter mile away. This stream enters the North Saskatchewan River, a major continental waterway of both aboriginal and historic times.
In Wisconsin current efforts are underway to archaeologically test the profiles of presumed petroforms in order to establish their authenticity. This is necessary since, without such effort, some of these boulder piles can not be immediately distinguished from the piles accumulating at the edges of cultivated fields by stone clearing. There are known cases too, where large circular formations were in fact, destroyed by land clearing and cultivation. This highlights the fragility of petroforms.
They are, in fact, the most fragile of all rock art. The mere kicking of a stone may destroy an astronomical alignment. And, field clearing, especially on hilltops may have destroyed hundreds, if not thousands, of such arrangements. Searching for intact specimens today seems best focused on hilltops covered with old hardwood growth, locations too steep to cultivate and showing long time natural integrity by virtue of very old forest. Such sites are now being discovered in the mid-continent both in Wisconsin and in Missouri, largely because there are committed investigators who know where to look!
In east-central Wisconsin a series of formal archaeological tests have been initiated at purported petroform sites. These efforts are being undertaken to resolve the problem of distinguishing between land clearance and aboriginal origin for boulder concentrations or alignments. While there are numerous intrinsic evidences for aboriginal construction, there remain lingering possibilities of non-Native origin. These latter prospects are almost always attributed to the removal of glacial erratics from land destined for cultivation. Extensive observations of cleared boulder concentrations reveal distinctive attributes almost never present among concentrations deemed to be aboriginal. Distinctions may be enumerated as follows:
1) Lines of boulders immediately adjacent to cultivated lands. These are usually within fence lines, since removing them to positions beyond the fence lines would usually entail their removal to a neighbor’s land.
2) Piling of boulders at the ends or sides of cleared land. Clearing lines are also piled. Cleared piles are characterized by extreme variability in size, from very large (.5 ton) to small, fist-sized cobbles.
3) Equipment marks left by bulldozers, front-end loaders, plows, etc. remain bright and are easily noticed on specimens in piles or piled lines.
4) Cleared boulders normally exhibit less lichenation, and far less recolonization of lichen than those of authentic Native construction. Low, tight forms of crustose lichen have extremely slow rates of growth. Earth, especially that containing high amounts of iron will show a boundary between original subsurface facets and those exposed to full lichen growth.
5) Random positioning of boulder lines from clearing will not usually exhibit significant astronomical orientations. Simple cardinal alignments are expected in these cases since these are lines often used in land surveys. However, if true cardinal lines are encountered, they may be aboriginal since these would not have employed a magnetic bearing – the usual way of lining up lands in early clearing. Magnetic directions may be as much as several degrees off “true.”
6) The verification of land use is vital. The exact chronology of clearing will eliminate some cases of boulder concentration from such clearing. These would be cases in which it can be positively shown that clearing took place after concentrations (with form, orientation, and other attributes) were known to have been in place.
7) Tree growth in the vicinity of boulder concentrations can aid in the testing, for historic or prehistoric origins. Ancient hardwood growth or such time-sensitive trees as cedars surrounding a potential feature may produce a clear case of pre-settlement activity. The absence of forest isolation may tend to confirm a land clearing origin, especially if taken along with other criteria.
8) A less thought of criterion has been suggested by Herman Bender. The original surface of erratics may have had highly conspicuous coloring and reflectance, and may have been selected for their visual nature in petroform construction. The exposure in stream cutting, for example, might have, at times, left them in pristine condition – having been protected in the morainal matrix.
9) Differential erosion between the surface of boulders above and below the soil line may be conspicuous. If a higher degree of surface erosion characterizes the above-ground part of a boulder, it may be concluded that it has been exposed longer than that protected in the ground. In some cases, windblown sand may abrade the exposed surface, making it far smoother than that below the ground.
10) The depth of a boulder may be significant in assessing the antiquity of its original positioning. Shallow depth may suggest a recent origin while deep positioning in the soil may suggest a longer period in the ground. If the petroform is linear or circular, and boulders appear to be missing, probing may be done at the expected locus of the “missing” boulders, as was successfully done by T.H. Lewis.
11) Usually, but not always, there is a marked variability in size for cleared boulders, while there is a tendency (very strong in some cases) for the boulders in petroforms to be uniform in size. Different sizes may characterize different sites. Theoretically, in the production of ritual features, youngsters may be involved, accounting for relatively small boulders being among those placed in the feature.
The archaeological testing of sites in east central Wisconsin has included all of the above assessments, as well as the introduction of one meter test squares at the edge of suspected formations. This is being done to evaluate the soil and/or cultural associations of the feature. In tests so far, suspected features are shown to be “pavements” – singly placed boulders, essentially without piling. These boulders are placed next to each other, at times with evidence of fitting. The vertical position in six tests is seen to have been placed upon an older soil surface some 12.0 cm below the modern one.
Cultural materials yielded in excavations profiling the edge of a feature do not necessarily establish the full cultural and geomorphological provenience of the feature. Placement on the same plane does not establish articulation either. At the Hutchings Petroform Site near the north shore of Buffalo Lake (a widening of the historic Fox River), cultural materials were encountered at levels 3, 4 and 5 in a unit profiling the west end of a perfectly arranged crescent oriented exactly to the summer solstice. The placement, however, seems to be slightly above the cultural materials, but resting upon an older soil zone. There is no doubt about the materials. There is a flake of Hixton Silicified Sandstone (Hixtonite) exclusive to Silver Mound in western Wisconsin, and entirely dependent upon human distribution. There is also a chalcedenous flint flake, and two fragments of fire- broken rock. Thin patches of fully disintegrated bone occurred at levels 4 and 5. The soil is highly acidic.
While proof of a relationship between the cultural levels and the petroforms are not established at the Hutchings Site, an aboriginal presence at the exact location is established. Moreover, one may speculate about that presence. The materials contain no pottery which is plentiful in local domestic sites having it. The use of Hixtonite is quite popular during preceramic times, and is common in Wisconsin during Paleo-Indian times. Despite the lack of a proven connection, the occasional inference of Archaic provenience for petroform sites in east-central Wisconsin (based upon a negative correlation with ceramic sites) is worth reflecting upon.
This latter observation applies particularly to another larger site being tested along the west branch of the Fond du Lac River in eastern Wisconsin. This is the Peachy Site which contains three large oval shaped pavements in an area not yielding pottery, but which does yield large quantities of diagnostic Archaic and Paleo-Indian lithics. Along with this is an array of Old Copper artifacts. The lithics present here form a strong representation of Old Copper-associated lithics; like Raddatz Side-Notched and Osceola projectile points, in classic form. The pavements here are dated to before clearing in both cases, and the shapes are possible animal representations, perhaps turtles. The boulders are fairly uniform and deeply imbedded in the soil. There are many evidences of prehistoric origin, strengthened by the placements being at least 20 cm below the surface and nearly to the underlying lacustrine clays. The potential timing for their placement could be Archaic, consistent with the bulk of the surface recoveries nearest to them.
One pavement at the Peachy Site is aligned at 123°, and has an easterly outlying boulder (also deeply embedded in the soil) through which the alignment is made. At the west end, no boulder is present, but the owners note that it was piled upon the pavement, having been relocated from the outlying position during clearing. There is another interesting attribute of the overall configuration. The long axis of the pavement exactly bisects the form, and a lateral line at the longitudinal center also bisects the form. Without reference to the other criteria this specimen (Feature #2) seems to exhibit more symmetry than would arise from pure coincidence. The 123° alignment is that of mid-winter sunrise, and the back azimuth is, of course, mid-summer sunset, far from that expected from Germanic farmers of the mid-nineteenth Century.
Perhaps the most compelling fact about the Peachy petroforms is that Mr. Peachy cleared the land near the petroforms, and in doing so destroyed a large circle of boulders some 40 feet in diameter. He pushed these boulders to a location well north of both pavements and they remain there beneath a 200 year old oak tree (there is a 300 year oak next to it!). Stone artifacts had been found near this tree prior to the displacement of the circle. And, in the area of the former circle, a number of perfect projectile points of Archaic type were collected from the newly cultivated land. The lithics found near the huge oak before the displacement were also Archaic pieces. The recoveries all remain in the Peachy Collection, and one magnificent side-notched projectile point of exotic stone nearly 25.0 cm in length was recovered from what would have been the center of the boulder circle. It is a classic Osceola point, fully consistent with a classic Old Copper “spud” found nearby.
About ten years ago, a Peachy family member, overcome by curiosity, used a front-end loader to expose the surface below one of the boulder pavements. Now referred to as Feature Four, the surface beneath the boulders exposed a small collection of weathered bone. Initial examination suggests that the bones represent at least two specimens of the dog family. They are especially interesting because the heads (including mandibles and all dentition) as well as the tails are missing. There are also butchering marks on some of the joint bones. Taken all together these facts point to dog ritual, a not uncommon practice among historic Indians of the sub-arctic. Clearly the animals had been skinned, with the tail and head “going with the hide.”
While only recently initiated, the application of clinical archaeological procedures in petroform investigations promises to greatly enhance our knowledge of the phenomenon. While advancing the processes of discrimination, this work will also shed light on the cultural and temporal provenience of petroforms, goals long thought to be largely unattainable. Page Up References
Behm, Jeffrey 1991 “The Fox Petroform (47-WP-214) Waupaca County, Wisconsin,” Fox Valley Archeology, Vol. 17, pp. 31-32, Oshkosh.
Behm, Jeffrey, Herman Bender, and Franklin Farvour 1989 “The Krug Petroform Site In Fond du Lac County, Wisconsin,” Fox Valley Archeology, Vol. 14, pp. 28-54, Oshkosh.
Danziger, E and R.T. Callaghan 1986“The Southeastern Manitoba Petroforms,” in Studies in Manitoba Rock Art: Petroforms, Province of Manitoba Historic Resources Branch, 2nd Edition, Second Printing, Winnipeg.
Lewis, T.H. 1889 “Stone Monuments in South Dakota,” American Anthropologist, Vol. 2, No. 3-A, pp. 159-167.
1890 “Stone Monuments in Northwestern Iowa, and Southwestern Minnesota,”
American Anthropologist, Vol. 3, pp. 269-279.
1891“Boulder Outlines In The Dakotas, Surveyed In The Summer of 1890,” American Anthropologist, Vol. 4, pp. 19-24.
Steinbring, Jack 1970“The Tie Creek Boulder Site In Southeastern Manitoba,” in Ten Thousand Years: Archaeology In Manitoba, Walter Hlady, Edit., pp. 223-268, Manitoba Archaeological Society, Winnipeg.
|