Monday, March 05, 2012

The Petroform Phenomenon of Southeastern Manitoba and Its Significance by Leo Pettipas

The Petroform Phenomenon of Southeastern Manitoba and Its Significance
Originally published 2004
Latest revision 30 October 2010

Leo Pettipas
Manitoba Archaeological Society

Petroforms are defined as cultural features formed on the ground surface “by the placement (not piling) of stones so as to create the outline of a figure” (Buchner and Callaghan 1980:72). Also known as “boulder mosaics” and “boulder outlines”, petroforms have been found in Minnesota, the plains states of North and South Dakota, Iowa, Nebraska and Montana, and in all three prairie provinces. As a group, they take the shape of animals, humans and geometrics, the latter including major components of the well-known “medicine wheels” of the western grasslands.

Probably the earliest reference anywhere to a petroform is to be found in the 1793 journal of Alexander Mackenzie. Referring to Portage du Bonnet on the lower Winnipeg River in what is now Manitoba, Mackenzie mentioned “a custom the Indians have of crowning stones, laid in a circle, on the highest rock in the portage, with wreaths of herbage and branches” (Lamb 1970:109-110). Scholarly references to non-Manitoba petroforms began to appear a century later (Lewis 1889; Thomas 1894), and in his presentation on such features originally published in 1913, F.W. Hodge included Manitoba in their overall range of distribution, quite possibly based on Mackenzie’s notation. Elsewhere, scholarly interest in them has persisted to varying degrees ever since. In Manitoba, research has been focused on the Whiteshell, since it had long been recognised that a large number of such sites was clustered in that area.

Researchers have sought answers to three basic questions concerning the petroforms: (1) how old are they, (2) who created them, and (3) why were they built? A fundamental, and logical, assumption from the outset has been that they were constructed by Aboriginals, and so in the early years of study, enquiries were directed toward Native informants. The answers received about certain western high plains features disclosed that they were monuments to specific events or prominent individuals (Hodge 1971:69; Kehoe nd:90), and some can be attributed to a particular nation (Simms 1903:374).

Early Petroform Studies in Manitoba
It was not always clear from the Aboriginal accounts just how old the features were or which people (tribal group) was responsible for their construction. However, Native testimony, in my opinion, has proven to be the most reliable source as to their age, authorship and function, at least in Manitoba. Most of the available cultural-historical information actually predates the advent of explicitly scientific, university-based field research by professional archaeologists, which did not begin until the late 1960s (Steinbring 1970:231). The earliest account of the Manitoba petroforms, other than that of Mackenzie in the late 18th Century, appeared in a newspaper article published in September of 1934 (Anonymous 1934a:1). A local homesteader guided a party from the Manitoba Museum to two sites in the Tie Creek-Nutimik Lake area of what is now Whiteshell Provincial Park. The group included a reporter from a Winnipeg newspaper, and the resulting article contained some very interesting and insightful information.

For one thing, reference was made to some unidentified “students of Indian history” who characterized one of the places, known in the recent literature as the Tie Creek Site (EaKv-1), as a “college” where novitiates of the Grand Medicine Society or Midewiwin (explicitly identified as such in the article) went to acquire “the lore which carried them through the four degrees of their craft” (Anonymous 1934a:1). It is not at all clear in reading the article how the un-named “students of Indian history” arrived at this conclusion, because the article states that “the Native Indians of today have apparently utterly forgotten” the site, implying that Aboriginals were contacted and interviewed, with negative results. If that is indeed the case, one wonders how the place could have been as confidently and explicitly attributed to the Ojibwas and the Midewiwin as it was in the article. Certainly the site’s declared functional attributes bespeak of a respectable grasp of Native culture as well as a capacity for insightful interpretation on someone’s part.

Nor is the reference to special teaching places unprecedented; it calls to mind a similar locale to the east that is alluded to in the early literature by the Ojibwa writer George Copway: “Then they [the party to be made medicine men and women] are taken to another place with our medicine men, and are taught the science of medicine. After receiving instructions, another day was allotted to give them instructions on morality” (Copway 1851:30). Much more recently, Ojibwa writer Basil Johnston (1995:24) has called attention to places where “the elders brought youths, especially chosen for their strength of character, to prepare and instruct them as future custodians and stewards of the spiritual and intellectual heritage of the Anishinaube [Ojibwa] people.”

These places, both in Ontario, were in remote locations and contained “teaching rocks” on which were inscribed petroglyphs (symbols pecked into the rock surface) or pictographs (rock paintings), neither of which has yet been found at Tie Creek nor anywhere else in the Whiteshell area of Manitoba.

In the meantime, the early newspaper article provided at least provisional answers, either directly or indirectly, to all three who-when-why questions. For one thing, it repeatedly attributes the Tie Creek petroforms to the Ojibwas. According to modern-day historians, the Ojibwas began arriving in eastern Manitoba in the latter half of the 18th Century along with the westerly-expanding fur trade (Hallowell 1967:112-113; Bishop 2002:81). In other words, the Tie Creek Site would be postcontact in age if, as the article asserts, its origin can be confidently traced to the Ojibwa and if the Ojibwa occupation of Manitoba does not pre-date the 18th Century. And finally, it was identified in the article as a place of instruction.

As an aside, it is interesting to note that the Manitoba discoveries were announced in a pair of Australian(!) newspapers (Anonymous 1934b; Anonymous 1935) shortly after they were reported in the Winnipeg Evening Tribune. Rather amusing is the fact that the Australian accounts compared/equated the Manitoba petroforms with Britain’s Stonehenge, even though the Tribune coverage was quite emphatic that the “crude designs” of the Canadian manifestations “don’t even suggest [the] architecture of Stonehenge.”

Douglas Leecham

In 1949, Dr. Douglas Leechman, head archaeologist with the National Museum of Canada in Ottawa, published a brief, one-page article in the Canadian Geographical Journal about the Whiteshell petroforms. He had either visited the Tie Creek Site or was otherwise familiar with its boulder alignments. He made one particularly astute observation in his paper that corresponds rather well with part of what appeared in the 1934 Winnipeg newspaper article. He wrote, “the arrangement of the boulders suggests that they may have been used in the religious ceremonies of the local Ojibwa Indians, since the outlines are similar to designs on birchbark rolls used in these rituals” (Leechman 1949:274). In 1955, Leechman had occasion to write a letter to Mr Hartwell Bowsfield, Secretary of the Manitoba Historic Sites Advisory Board, concerning the petroforms. The letter reads, in part, “The biggest [feature at Tie Creek] in its outline resembles native drawings on birch bark of The Midewiwin Lodge, a sort of secret society which had various initiations and degrees.”
Leechman’s comparison of the biggest feature at Tie Creek (“Feature 1”; see Fig. 1, this paper) with the floor plan of a Midé lodge (Midéwigan) may have been more insightful than even he himself realised. The outside periphery of both the Tie Creek feature and the Midé lodge describes an elongated rectangle; but in addition, more than one description of the lodge notes a line of stones running down the middle. J.J. Hargrave’s account of a Midéwigan erected near Lower Fort Garry (now Winnipeg) in the 1860s includes mention of a line of poles running lengthwise through the centre “with large stones at their bases” (Hargrave 1871:197, emphasis mine).

Similarly, Sister M. Inez Hilger (1951:56) noted “a row of 20 stones, each about 6 inches in diameter set about 4 feet apart” running down the middle of a Lac Court Oreille Midéwigan that saw use in 1935. The big rectangular feature at Tie Creek likewise includes such a central line of stones, although a gap exists in the centre of it. Leechman’s correlation of the Tie Creek feature with the Ojibwa in particular is also significant. As it turned out, his comparison of it to the etchings on the birchbark scrolls had important implications, as subsequent enquiries into the Tie Creek Site were to show.

Cecil Patterson

The same year that Leechman published his Canadian Geographical Journal article, Mr Cecil Patterson, a forest ranger with the Manitoba Department of Mines and Natural Resources, noticed the Tie Creek Site while on an aerial fire patrol. Patterson had a personal interest of his own in regional Native heritage, and this, along with encouragement by Leechman, prompted him to seek information about the place. In 1952 he sketched a floor-plan of the entire site (Fig. 2) and presented it to Chief George Barker of the Hollow Water Band, asking him to enquire within his community about its significance. Mr Joe Black, reputed to be around 100 years old, had been a visitor to the Tie Creek Site in his younger years. He testified that the medicine men of all degrees and their apprentices, from all the bands around the country, went to the place once a year and “got ‘medicine’ out of the stones. They opened the larger stones and got out ‘medicine’.” After these matters had been attended to, they returned to their respective communities and held the Medicine Dance in which all the band members participated (Wells 1952; Hall 1960:16).

This functional explanation of the Tie Creek Site differs from the interpretation published in the 1934 Winnipeg Evening Tribune article, but it should not be concluded that it (the newspaper article) necessarily covers the full range of activities conducted at Tie Creek. Similarly, Joe Black’s information may not tell the entire story. The 1934 Winnipeg newspaper article does, however, make clear reference to the use of the place by the medicine people and a connection with the Midewiwin. In this particular regard, the earlier published statements in the newspaper article and that by Leechman are noteworthy.

In addition, it should be pointed out that the concept of acquiring medicine from stones is found elsewhere in the published literature. Irving Hallowell relates the story of Yellow Legs, a renowned Lake Winnipeg Midé who had a special boulder that possessed contours suggestive of eyes and a mouth. It was used in the Midewiwin for many years. When Yellow Legs tapped this stone with a knife, it would open its mouth, whereupon he “would insert his fingers and take out a small leather sack with medicine in it” (Hallowell 1975:148). The medicine “would be made into a concoction, which was then shared by all present” (Hallowell 1936:48).

Further Native Testimony

In addition to the large and complex Tie Creek Site, the Whiteshell contains a number of smaller loci with one or more petroforms. These comprise animal effigies, principally snakes and/or turtles, or geometrics such as squares, lines and ellipses. The smaller geometrics continue to be problematic in terms of their function, as no Aboriginal testimonies regarding them are available. The animal forms, on the other hand, have been placed in behavioral context by none other than George Barker, the same individual who proved so instrumental in obtaining from Joe Black the statement on the significance of the Tie Creek Site. In 1979, Mr Barker published an autobiography that included a chapter entitled “Medicine Dance”. His brief description bears out the role of certain turtle and snake petroforms:

A row of stones, placed from smallest to largest, was often used to aid the medicine man. These formed snakes of various lengths. The sick person was laid beside the snake, which would then begin to move. Often, this resulted in a cure. The stones also sometimes became turtles (Barker 1979:99).

Again, this report of stones assuming animate properties under special circumstances has counterparts in the literature. Hallowell (1975:148) cites an Ojibwa informant whose father was leading a Midewiwin ceremony in which a large round stone was present. The Midé

got up and walked around the path once or twice. Coming back to his place he began to sing. The stone began to move “following the trail of the old man around the tent, rolling over and over, I saw it happen several times and others saw it also.” The animate behaviour of a tone under these circumstances was considered to be a demonstration of magic power on the part of the Midé.

Formal Scientific Research

As was noted above, personnel connected with The Manitoba Museum were showing interest in the petroforms as far back as 1934, and they continued to do so over the following three decades mainly by facilitating visiting researchers, in particular Richard S. MacNeish and Boyd Wettlaufer of the National Museum of Canada, in their studies. In the late 1950s and early 1960s, however, Richard Sutton (1965), also of The Manitoba Museum, prepared a series of field grid drawings and eventually published them in a brief paper in the Manitoba Archaeological Newsletter. This work was soon followed by university-based research that began in 1967 and continued for more than two decades.

The lead authority in the university research was Dr Jack Steinbring of the University of Winnipeg. Dr Steinbring was aware of the earlier professional and non-professional enquiries into the significance of the petroforms, but he encountered what he perceived as shortcomings in their work when measured against late 1960s professional research standards. In 1970, he wrote:

On June 18th, 1960, Mr. Frank Hall presented a paper on boulder sites at the opening of the Nutimik Museum in the Whiteshell. Mr. Hall ... gave what appeared to be a confirmed cultural association through data from an Ojibwa informant. Mr. Hall reported that Old Joe Black of the Hole River Band (Wanipigow) had explained, through an interpreter, the use of “mosaics”in Ojibwa magico-medicinal activities. On further enquiry we learn that Mr. C. Patterson had collected, through an interpreter, some data from Mr. Black on the MITAYWIN [sic](or Grand Medicine Society). This was in 1952, only a year or so before Joe Black died at the age of 100 (Steinbring 1970:231).

The foregoing paragraph is essentially in keeping with what has been stated in the preceding pages, with one exception: it notes that Mr Patterson had collected data on the Midewiwin. This is true, but Patterson’s principal focus from the outset was on the petroforms, not the Midewiwin. As it turned out, the latter was of direct relevance to the story behind the boulder formations, which came to light thanks to Patterson’s interest in the rock formations.

Dr Steinbring goes on to state his reservations about the non-professional findings:

It is not certain, because of the problems of translation, if Mr. Black was referring to the common placement of three “medicine stones” in MITAYWIN and other Ojibwa medicine ritual, or if he was referring to actual effigy arrangements. The latter seems very doubtful, and it is certain that his remarks pertained to the Hole River area only, and probably to Black Island itself where the MITAYWIN was last celebrated in 1926 (ibid.).

Dr Steinbring’s comments are worthy of further scrutiny on several counts. First, he raises the question of “the problems of translation.” Unfortunately he does not elaborate in his publication on what exactly these problems were and why they should have had a critical bearing on the outcome of Patterson’s research. Second, he is unsure if Mr Black was referring to “three medicine stones” rather than petroforms. From the government documents it is clear (to me) that Mr Black was indeed referring specifically to the Tie Creek petroform site because he was responding to the plan of that site drawn by Cecil Patterson. I do not at all consider “very doubtful” the idea that Joe Black was referring to the petroforms.

Third, although Dr Steinbring is certain that Mr Black was alluding to the immediate Hole River area, and Black Island in particular, I submit that his certainty is ill-founded. Again, based on the Tie Creek floor plan drawn by Patterson and used by George Barker in his enquiries of Mr Black, it seems quite certain to me that the latter was referring explicitly to the Tie Creek Site, which is 140 kilometres from the Black Island/Hole River area.

In the event, archaeologists devised or adopted various ways of their own of eliciting information about the petroforms. Astronomical observations, typological comparisons, erosional facet measurements, lichenometry and boulder numbers analysis are all cited as scientific means of studying the features. However, successful application of these techniques has proven highly problematic, as will be shown below, and the actual documentation arising from Patterson’s and Barker’s enquiries lay buried in government files and did not come to the attention of the academic community until the late 1980s. In the meantime, the archaeologists were left to their own devices in their efforts to determine the age, authorship and function(s) of the Whiteshell petroforms.


To explore the antiquity of sites, archaeologists employ two standard techniques – radiocarbon assay and cross-dating. The radiocarbon, or C-14, method requires the remains of once-living entities (plants or animals) such as bones or charcoal -- the kinds of refuse one might find in a former habitation site wherein a range of organic by-products of everyday living were produced. Unfortunately, the petroform sites have consistently proven to be clean of such material, and so application of the C-14 technique has not been possible.

Cross-dating involves the comparison of artifacts from an undated site with those from others which have already been C-14 dated. If the two sites both contain stylistically similar artifacts, it is concluded that the undated site is comparable in age to the C-14 dated one. Again, such an approach is frustrated at the petroform sites, where artifacts other than the rocks themselves are absent. Hence, no artifactual comparison is possible. A variation on the cross-dating approach is site proximity analysis. This simply involves locating and dating habitation sites occurring in the same area as the petroform, and assuming that they were created by the same people who built the nearby petroform feature.

However, in an area like southeastern Manitoba where people have been living for some 8000 years, the connection between any two adjacent sites, only one of which can be independently dated by C-14, may be more apparent than real. Proximity in and of itself is no guarantee of connectedness or relationship, and the use of it as a working assumption “involves questionable reasoning” (Buchner 1980:96).

The one dating technique that theoretically can be conducted at petroform sites alone is lichenometry. This method is based on the rate of lichen growth on rock. If one can monitor the growth rate of a newly-introduced patch of lichen over several years, it should be possible to estimate the age of established lichens growing on and around an adjacent petroform. For this to work, it has to be assumed that the existing lichen associated with the petroform is the direct “descendant” of the original growth, and that there has been no interruption in the history of growth via fire or disease. It must also be assumed that there was no delay in the commencement of growth of the original patch, and that there was no old growth already on the rock before it was incorporated into the petroform.

These variables are difficult if not impossible to control for, and to demonstrate how tenuous lichenometry can be, one researcher notes that “features only a few decades old may easily acquire the lichen and moss growth usually associated with prehistoric features, and of course, stones with ancient lichen colonies on them may be used in construction of new features” (Buchner1992:67). This is precisely what one archaeological team observed in the case of on a 1937 surveyor’s cairn in the Whiteshell: the amount of lichen growth on the cairn “would make it as old as any petroform” yet found in the area (Danziger and Callaghan 1986:152).

A final avenue in seeking the age of the petroforms is through interviewing Native informants, a procedure that is more ethnographic than archaeological. This may or may not be productive, for a variety of reasons. For one thing, Indigenous people tend not to gauge antiquity in the Western fashion. As Hallowell (1937:667) noted, “one hundred and fifty years is the outside limit of any genuine past” among the Lake Winnipeg Ojibwa with whom he was familiar.

Events attributed to so distant a past that they cannot be connected with any known generation of human individuals are simply described as having taken place “long ago.” Consequently we are plunged into a bottomless mythological epoch that lacks temporal guide posts of any conventional sort (ibid.).

This state of affairs is incompatible with the objectives of archaeologists who seek to express the antiquity of heritage features in terms of years BP, BCE, CE, BC and AD. Furthermore, if the present-day Natives of an area are not direct descendants of those who lived there in antiquity, it may be unreasonable to expect them to possess traditional knowledge of the genesis of local heritage features. And even if they are, many or most may disclaim any knowledge of them, either because such knowledge has been lost or else because the informants are hesitant to talk about them to outsiders. The situation has been summarized rather well by the Three Fires Society (1990:iii-iv):

Although scholars have attempted to interview Native people in nearby communities about their interpretations of the petroforms, such information has not readily been forthcoming to date. This is not surprising, given the well-documented history of repression of Native culture and religion in North America. In addition, such information must be “earned” in traditional ways, not through interviews. Some of the knowledge is also only available to the formally taught and initiated. It is thus not available to either the Native or non-Native general public, nor even to the total membership of the Midewiwin.

Under these circumstances, it is not difficult to understand why academic research “couldn’t establish a connection between the [Tie Creek] site and existing cultures” (Steinbring 2005). When R.S. MacNeish published his cultural chronology for southeastern Manitoba in 1958, the petroforms were not included in any of the constituent foci. This is because the stone formations did not lend themselves to dating according to the relative or absolute chronological techniques available at the time. If, as is contended here, the petroforms are indeed an entirely postcontact phenomenon, then MacNeish was wise, as it turns out, not to assume a place for them in a precontact cultural chronology.

It is surprising that the avocational efforts of Mr Patterson and his associates in the 1950s, and other “students of Indian history” some two decades previous, seem to have, inadvertently, dated the Tie Creek site by virtue of its association with Ojibwa culture. Interestingly, after more than ten years of research into various aspects of the petroforms, the prevailing opinion within the scholarly community was that there was no evidence of an explicitly Ojibwa connection (Steinbring 1980:311). “Despite concerted attempts in southern Manitoba, no Native informants claiming knowledge of the meaning of these structures have been found” (Buchner 1980:96); “a number of elderly Native informants were taken to the sites, but none had any notion of the function or age of the features. This suggested [to the archaeologists] that most, if not all, petroforms were of prehistoric age” (Buchner and Callaghan 1980:73; see also Danziger and Callaghan 1986:61).

The “negative evidence” encountered by Dr Steinbring, then, seems to have deflected him and his students away from the prospect of a postcontact Ojibwa authorship, leading them to place “the initial date of [petroform]construction between 500 B.C. and A.D. 1” (Buchner 1976:11), and hence pre-Ojibwa. With little else to go on, the archaeologists resorted to the above-mentioned proximity analysis approach and hypothesized that some components of the Tie Creek petroforms are upwards of 3000 years old, while the snakes and turtles date to “perhaps around A.D. 1200” (Steinbring 1985/1993:26), again, making them too early to have been built by the Ojibwa. These assertions were still in vogue after the turn of the 21st Century: in 2003, it was believed that the Tie Creek petroforms “were laid out on the open rock formations, at least 2,000 years ago, some of them perhaps 1,000 years earlier” (Steinbring 2003).

Since, as noted above, more reliable archaeological dating techniques cannot be applied to petroforms, nothing more conclusive was ever determined via professional research as to their antiquity, nor was the hypothesis of precontact origins challenged in print for many years after it was originally proposed. It was not until 1987 that discussions between a Native informant and a professional anthropologist yielded positive results regarding the age of the Manitoba petroforms. In September of that year, The Manitoba Museum’s Curator of Native Ethnology was among a party that visited the Tie Creek Site with the late Herman Atkinson, an Ojibwa elder, healer, Thirst Dance ritualist and Second Degree Midé. Mr Atkinson had personal, first-hand knowledge of the site, having visited it in the 1940s. He stated that the petroforms had been put in place “after the Whitemen had arrived” (Pettipas 1990:84), a confirmation of Joe Black’s affiliation of it with the postcontact Ojibwas and, by extension, the Midewiwin.


The aforementioned difficulty experienced by professional archaeologists in obtaining information from living informants likewise frustrated their ability to identify who it was that created the petroforms. They were forced to rely on generalized historical models, notably the “long series of westward population expansions” and the “steady stream of influences [that] proceeded along the major waterways from Lake Superior all the way to Alberta” (Steinbring1976:6-7) over the past three millennia. This scenario is taken to explain the existing continental distribution of the Algonquian language family, leaving one to conclude that it was Algonquians who created the petroforms since the latter are positioned geographically within the Algonquian east-west migration corridor.

Following, and being confined to, this line of reasoning, it was not possible for the scholars to be more specific as to who within the Algonquian language family might have produced the petroforms. By the late 1980s, however, the academic community appeared to be cautiously venturing a tribal affiliation for the Tie Creek Site: “While no exact connection with any specific native group can be established, the best evidence points to a link with the remote ancestors of the Ojibwa (Saulteaux) who occupy the reserves nearest the site” (Steinbring 1988:6). However, the jury was still out as of 2003; Dr Steinbring seems to have remained convinced at that time that “no connection to any specific tribal group has ever been fully established.” This opinion was re-affirmed two years later, when, in an interview (Montag 2005), he is reported to have said, “The [Tie Creek] images were commonly thought to have been created by the Ojibway. But we found evidence that the site was older than that.” Unfortunately, he does not reveal in this 2005 statement what form that evidence took, or just when it had come to light.

One scholar has advised that allowance should be made for the prospect that the northern Cree were somehow connected with the petroforms (Lanteigne 1990:127). Both archaeology and historiography have brought forward good indications that Cree-speakers inhabited southeastern Manitoba during the late precontact and early postcontact eras, as did the Assiniboin (Hallowell 1967:114-115). At the same time, there is not one scintilla of substantive evidence that the Whiteshell petroforms were routinely used, never mind created, by either Cree or Assiniboin people.


For as long as the Tie Creek petroforms have been studied by heritage aficionados and scholars alike, the assumption has been that they were of ritualistic and ceremonial significance. Native testimony has borne this out, but Native testimony alone has been able to provide particulars as to their function. Interested parties who were not privy to the testimony of Native insiders were relegated to offering speculative interpretations, most of which are highly tentative because a petroform site does not contain explicit evidence of esoteric behaviour or the ideology, intellectual underpinning, or worldview that lies behind it. The conundrum faced by archaeologists everywhere was poignantly expressed by Professor Grahame Clark in his thinking about the “megalithic” (big stone) sites of Europe:

Even if we could visit Bronze Age Britain and study at first hand the rites and practices associated with them, it would still be difficult to comprehend their underlying meaning: to probe the innermost consciousness of men who lived thousands of years ago by measuring and classifying stones, however meticulously, is manifestly vain (quoted in Hadingham 1978:224).

Thus, when Dr Steinbring (quoted in McConnell 1985) eventually suggested that the unique, 9-acre Tie Creek petroform site was for use by “perhaps some sort of medicine society”, he must have been basing his statement on something other than what can be revealed through the application of the archaeological field methods noted and discussed above.

In their efforts to discern patterns of past human behaviour, archaeologists seek out correlations between cultural and natural phenomena. Some good examples of this can be found in the conduct of “archaeoastronomy.” In the case of the petroforms, a consistency of alignment was observed among features of similar form. “Linear and elliptical arrangements are oriented along a northeast-to-southwest axis, such that at midsummer they closely approximate the rising azimuth of the sun.” Turtle effigies, on the other hand, “exhibit a 90% consistency of orientation towards the west” (Buchner 1976:11-12).

These and similar correlations have led to the suggestion that “the builders of these sites were not only cognizant of, but preoccupied with, the cardinal directions and the movements of celestial bodies” (ibid.). Unfortunately, if standard archaeological thinking is all we have to go on, this kind of general statement is about the best that can be hoped for; the attendant belief system, world view and ritual that went with such observances are destined to be, to use a hackneyed phrase, forever shrouded in mystery – assuming that the petroforms date to long-ago precontact times.

Likewise, there has been a noticeable correlation between snake and turtle effigies and certain geographical situations that has given rise to the hypothesis that the features are “trail markers” or “signposts” at portages between ricing lakes. This may seem like a perfectly logical conclusion to someone who is a product of 20th Century Western society and culture, but whether it actually corresponds to the motivations of past Indigenous people with a profoundly different way of life is very much open to question. In fact, if we accept George Barker’s explanation of them, the snake and turtle petroform sites were healing places, and their geographical locations may have been secondary if not incidental to their primary purpose.

Another means used by archaeologists to interpret their findings is formal comparison. This involves seeking out similarities in content, form or style between separate cultural expressions and, where such similarities are found, concluding that there was some manner of historical relationship between them. This modus operandi is focal to the “diffusionist” school of thought, which posits, for example, that Mesoamerican civilization was influenced by cultures in Southeast Asia, and that the Peterborough petroglyphs in Ontario are of Scandanavian Bronze Age cultural origin (Heine-Geldern 1959; Kelley 1990:3,4). When Leechman called attention to the similarities in form between the outline of Feature 1 at Tie Creek and that of the midéwigan on birchbark scrolls, there is not much question that he was inferring some manner of cultural-historical connection or relationship between the petroform and the Midewiwin. Dr Steinbring (1976:5) noted that turtle and snake motifs are common both to the known sample of Whiteshell petroforms and the mythology of the Ojibwa. However, Leechman’s suspicions drew independent support from the testimony of Joe Black, whereas Dr Steinbring has been indisposed toward an Ojibwa affiliation by virtue of the negative evidence he was left with from his own enquiries among Ojibwa informants.

A further difficulty with formal comparison is the amount of variation different researchers may allow for in arguing that the compared items are, or are not, meaningfully similar. It is often a matter of opinion as to what is “similar” and what is not. For example, Dr Steinbring (1970:243-244) has noted that a human-shaped petroform near Consort, Alberta (Fig. 3) in its general rectangular shape, length and possession of a “heartline”, is “reminiscent” of Feature 4 at Tie Creek (Fig. 4). The implications are that these and similar observations bear witness to “the westerly movement of Algonkian peoples”, and further that “the boulder phenomena of the Whiteshell form a base stratum from which a general western diffusion finally achieves a fairly broad Plains distribution” (Steinbring 1970:247).

However, another observer might not regard these perceived commonalities of form as being close at all, and argue that some other interpretation is in order. Either way, analysis of this sort is impressionistic and subjective, and is basically a variation on the “Diffusion vs Independent Invention” debate that typically ends with the two sides simply agreeing to disagree.

Discussion and Conclusions

Archaeology is generally agreed to be a scientific discipline. As such, it involves a process that goes something like this: the researcher makes observations, on the basis of which he constructs a hypothesis. The hypothesis is then tested to see if it is supported by additional facts. If it is, it is regarded as a good hypothesis and a credible explanation of whatever it is the researcher is trying to explain. It would be a real help if one could be assured that he will have access to all the facts relevant to the hypothesis. Unfortunately, in archaeology it is impossible to know when “all the facts are in”, simply because we do not know what the full range of relevant data looks like to begin with. That being the case, it is not possible to draw final conclusions about the past via archaeology: allowance must be made for the prospect that there are other data that have not yet been found – and indeed, may never be found because they have been irretrievably destroyed.

Thus, a “good” hypothesis is not one that has been proven conclusively to be true. Rather, it is one that has repeatedly been shown not to be false. The more and better tests that can be run, the more confidence we will have in the hypothesis. Unfortunately, petroform sites, as shown above, do not lend themselves to the testing of hypotheses using standard methods. Thus, one can hypothesize that the Tie Creek Site is 3000 years old based on occupation site proximity analysis. But since petroform sites are not subject to radiocarbon assay or cross-dating, and the lichenometric technique is currently fraught with difficulties, the original hypothesis cannot be tested, and hence we have no good way of knowing whether or not the hypothesis of a 3,000-year antiquity for the complex is a good one.

In sum, trying to fathom the cultural-historical significance of the petroforms without recourse to informants who were actually involved with them, and who are prepared to discuss them, is an exercise in futility. Without a doubt, the chief contribution made by professional archaeology to the study of petroforms is the ample series of written descriptions, photographs and diagrams of them produced by practitioners in the discipline (Steinbring and Muller 2010:2). But the interpretation of them as precontact by several hundreds and thousands of years is, in the present state of scientific research, highly questionable since the weight of tangible evidence, scant though it is, has pointed all along to an Ojibwa authorship that would be only a few hundred years old at most.

It was indeed fortuitous, then, that at least some information about certain types of petroforms has been obtained from Native informants, two of whom were or had been Mideg. However, the available information is minimal and sketchy at best, and it may remain so for a long time: as pointed out by the Three Fires Society, members of the contemporary Midewiwin are not well disposed toward the enquiries of social scientists, and so the prospects of additional testimony becoming available to the larger community are therefore remote.

It is concluded here, based on informant testimony, that the Tie Creek, snake and turtle petroform sites of the Whiteshell district of southeastern Manitoba were the creations of the postcontact Ojibwa, not “prehistoric” Algonquians, and that their functions were part of the Midewiwin religion. There is no reason at this time to think that they are ancient precontact features that were discovered by the Ojibwas in very late precontact or early postcontact times upon their arrival in Manitoba, and subsequently incorporated them into their culture. Rather, I hypothesize that they were “tailor-made” infrastructure created by Ojibwas expressly for use in Midewiwin instructional and healing contexts.

The prime raison d’être of the Midewiwin was the maintenance of good health and the combating of disease (Vecsey 1983:179), and the notable concentration of petroforms in the Whiteshell suggests that the area was a major centre for the conduct of the healing arts among the regional Ojibwa. Conventional wisdom has it that the geographic name “Whiteshell” derives from the megis shell, an integral component of Midewiwin tradition, symbolism and ritual.


Anonymous, 1934a, Unique Out-Door Cathedral of Ojibwas Believed Found. The Winnipeg Evening Tribune, 8 September edition.

Anonymous, 1934b, New Stonehenge: Great Stones in Canada Symbols of the Indians. The Canberra Times, 22 October, p. 2.

Anonymous, 1935, Canadian Stonehenge Relic of a Primitive Cult. Western Argus, 19 February, p. 35. Kalgoorlie, WA.

Barker, G., 1979, Forty Years a Chief. Peguis Publishers Ltd. Winnipeg.

Bishop, C., 2002, Northern Ojibwa Emergence: The Migration. In Papers of the Thirty Third Algonquian Conference, ed. by H.C. Wolfart, pp. 13-109. Winnipeg.

Buchner, A., 1976, The 1973 Rock Art Studies at the Tie Creek Site: A Preliminary Report. Papers in Manitoba Archaeology Miscellaneous Papers 2:10-22. Manitoba Tourism, Recreation and Cultural Affairs. Winnipeg.

1980, Archaeo-Astronomical Investigations of the Petroform Phenomenon of Southeastern Manitoba. In Directions in Manitoba Prehistory: Papers in Honour of Chris Vickers, ed. by L. Pettipas, pp. 89-109. Winnipeg.

Copway, G., 1851, Recollections of a Forest Life: Or, the Life and Travels of Kah-Ge-Ga-Gah-Bowh, or, George Copway, Chief of the Ojibway Nation. C. Gilpin. London.

Danziger, E. and R. Callaghan, 1986, The Southeastern Manitoba Petroforms: Assessment of an Expanded Sample. Papers in Manitoba Archaeology, Miscellaneous Paper 14:57-173. Manitoba Culture, Heritage and Recreation. Winnipeg.

Hadingham, E., 1978, Circles and Standing Stones. Sphere Books. London.

Hall, F., 1960, Medicine on the Rocks: The Strange Tale of Ojibwa Mosaics. The Bison 13(2):14-16.

Hallowell, A.I., 1936, The Passing of the Midewiwin in the Lake Winnipeg Region. American Anthropologist 38:32-51.

-----, 1937, Temporal Orientation in Western Civilization and in a Preliterate
Society. American Anthropologist 39:647-670.

-----, 1967, The Northern Ojibwa. In Culture and Experience, pp. 112-124. Schocken Books. New York.

-----, 1975, Ojibwa Ontology, Behavior, and World View. In Teachings from the American Earth: Indian Religion and Philosophy, ed. by Dennis Tedlock and Barbara Tedlock, pp. 141-178. George J. McLeod Ltd. Toronto

Hargrave, J., 1871, Red River. John Lovell. Montreal.

Heine-Geldern, R., 1959, Chinese Influences in Mexico and Central America: The Taijin Style of Mexico and the Marble Vases from Honduras. Actas del 33ra Congreso Internacional de Americanistas, San José, Costa Rica, pp. 195-206. San José.

Hilger, M., 1951, Chippewa Child Life and Its Cultural Background. Bureau of American Ethnology Bulletin 146. Washington.

Hodge, F.W., 1971, Handbook of Indians of Canada. Geographic Board, Canada. Coles Publishing. Toronto.

Johnston, B., 1976, Ojibway Heritage: The Ceremonies, Rituals, Songs, Dances, Prayers and Legends of the Ojibway. McClelland and Stewart. Toronto.

-----, 1995, The Manitous: The Spiritual World of the Ojibway. Key Porter Books. Toronto.

Kehoe, A., nd, Hunters of the Buried Years. School Aids & Text Book Publishing Co. Ltd. Regina.

Kelley, D.H., 1990, Proto-Tifinagh and Proto-Ogham in the Americas. The Review of Archaeology 11(1):1-10.

Lamb, W., 1970, The Journals and Letters of Sir Alexander Mackenzie. Cambridge University Press. London.

Lanteigne, M., 1990, Review of “Tie Creek Study: An Anishinabe Understanding of the Petroforms in Whiteshell Provincial Park.” Roseau River Chapter, Three Fires Society. The Rock Art Quarterly 1(3-4):126-129.

Leechman, D., 1949, Turtle Mosaic. Canadian Geographical Journal 39(6):274-275.

Lewis, T., 1889, Stone Monuments in South Dakota. The American Anthropologist 2(3-A):159-167. Washington, DC.

MacNeish, R.S., 1958, An Introduction to the Archaeology of Southeast Manitoba. National Museum of Canada Bulletin 157. Ottawa.

McConnell, W., 1985, Whiteshell Petroforms Site of Indian Rituals. Winnipeg Free Press, Saturday, 15 June. Winnipeg.

Montag, T., 2005, 2005 ESRARA Rock Art Conference: An Interview with Jack Steinbring. The%Middlewesterner%202005%20ESRARA.

Pettipas, K., 1990, Part 2: A Commentary. In Tie Creek Study: An Anishinabe Understanding of the Petroforms. Roseau River Chapter, Three Fires Society. Ginew, Manitoba.

Simms, C., 1903, Anthropologic Miscellanea. The American Anthropologist 5:374-375. Washington, DC.

Steinbring, J.H., 1970, The Tie Creek Boulder Site of Southeastern Manitoba. In Ten Thousand Years: Archaeology in Manitoba, ed. by W.M. Hlady, pp. 223-268. Manitoba Archaeological Society. Winnipeg.

-----, 1976, Boulder Art in Eastern Manitoba. Papers in Manitoba Archaeology Miscellaneous Papers 2:3-9. Manitoba Tourism, Recreation and Cultural Affairs. Winnipeg.

-----, 1985/1993, Dating Rock Art in the Northern Midcontinent. Paper presented at the Twelfth Annual American Rock Art Research Association Conference, Santa Barbara, California, 1985; published in American Indian Rock Art XII:15-29, 1993.

-----, 1988, Quarrying. Winnipeg Free Press, 19 February, p. 6.

-----, 2003, Current Research on Wisconsin Petroforms. %20research.htmhttp://

----- and N. Muller, 2010, North American Petroforms: Questions of Their Chronological and Cultural Placement. IFRAO Congress on Pleistocene Art of the World, September 6-11, 2010, Ariège-Pyrénées, France; Symposium: Pleistocene Art of the Americas (Pre-Acts), pp. 1-6

Sutton, R., 1965, The Whiteshell Boulder Mosaics. Manitoba Archaeological Newsletter 2(1):2-11. Winnipeg.

Thomas, C., 1894, Report on the Mound Explorations of the Bureau of Ethnology. Twelfth Annual Report of the Bureau of Ethnology 1890-1891, pp. 35-43, 534-536. Washington, DC.

Vecsey, C., 1983, Traditional Ojibwa Religion and Its Historical Changes. American Philosophical Society Memoir 152. Philadelphia.

Wells, H., 1952, Reference to the Boulder Mosaics near Sturgeon Falls in the Whiteshell. Ms, Manitoba Department of Mines and Resources. Winnipeg.


pwax said...

I have an almost exclusively positive reaction to everything the author says in this article, especially what he says about methodology.

One sour note though was this:
"In sum, trying to fathom the cultural-historical significance of the petroforms without recourse to informants who were actually involved with them, and who are prepared to discuss them, is an exercise in futility."

That does not mean the continued study of the physical data is pointless and, for my money, some speculation is needed to advance even simple the simplest observation - let alone full hypotheses.

pwax said...

From an email exchange bewteen PWAX and the author LP:

[PWAX]Your article is thought provoking. I particularly like your methodology but when you say it is "futile" to look for a deeper understanding of the cultural meaning, I am not sure that is quite fair. Some kind of speculation plays an important role in forming hypotheses and in making physical observations. Also, I don't think the cultural meaning is the only end of the study and that there is plenty to pursue at the physical level - especially in New England.

It is helpful to me to hear that snakes and turtles had particular meaning in the healing and medicine of the Indians and it is interesting that the symbols are similar to what appears on birch bark. It is hard to believe the Indians would have invented the symbology post contact.

[LP]And indeed, the symbology may be precontact and pre-Midiwewin -- that would be worth looking into. However, my point is that the use of boulders per se ties in expressly with the Midewiwin, which is a post contact phenomenon. If someone could somehow demonstrate that the Midewiwin is precontact, then it would be a whole new ball game.

[PWAX]So why did they start using boulders to represent it?

[LP]Good question; my only answer is ... why not? For centuries, European artists (sculptors) have used rock (marble) to create statues, even they had other media as well with which to create their art. Maybe rock/stone had special, spiritual meaning to the Natives --- I give a few (post contact) examples of this in my paper.

[PWAX]Was there a lack of birch trees?

[LP]Not likely -- this question could be answered by checking the regional pollen diagrams, but offhand I'd say that birch trees go back a long ways hereabouts. So I doubt that it would have been a lack of birch bark that motivated the construction of petroforms.

[PWAX]Did they need large structures only post contact?

[LP]Hard to say. Certainly the Mayas built large structures in precontact times, not to mention the Britons' Stonehenge, which is judged to be "prehistoric."

[PWAX] As I said - thought provoking

pwax said...

I can just imagine Pettipas and Steinbring arguing over the age of these sites. It comes down to this: Pettipas found an Indian who claimed knowledge, Steinbring did not. The Indian may or may not have been telling the truth. From my perspective you should not rely on word of mouth alone.

pwax said...

Writing in 2017:

Pettipas says:"trying to fathom the cultural-historical significance of the petroforms without recourse to informants who were actually involved with them, and who are prepared to discuss them, is an exercise in futility. "

This may be correct in the cast of petroform sites which are extremely rare. But in the case of rock piles sites in MA, there are thousands of sites - permitting hypothesis testing at a grander and more statistical scale than is being dismissed by Pettipas.