A blog by our founder, Ron Williamson
I now spend some of my time publishing the results of past ASI projects with which I was involved, a task consistent with one of the core values of the company, which is to disseminate the results of our work. In this case I was also motivated by Order of Canada recipient and former Chief of the Mississaugas of Credit First Nation Carolyn King, who has argued for decades that the small archaeological sites of her ancestors are as significant as the large settlements of other Indigenous peoples or settlers. This is clearly the case for the Graham site (AkGx-41), which was located in the Town of Caledon (Figure 1). An article on the site has just been published by me in Ontario Archaeology.
The site was found in 2002 during pedestrian survey of a cultivated field on a slight knoll adjacent to a tributary of the Credit River. Further investigation of the location the following year revealed a small ceremonial site with one feature containing over 250 pieces of animal bone that were burnt to the point they had become white and highly fragmented. Another feature originally contained a cache of thermally fractured Meadowood-like bifaces -a type of double-sided stone tool-, along with over one hundred copper beads and other copper tools.
With only four features and no evidence of habitation, but with indications of a feast and the “ritual killing” of associated items, this three-thousand-year-old site seems to represent a place of ceremony, perhaps for a family. This family, in turn, might have been part of a larger macroband that gathered at a different time of year. It is known that for the historic Anishinaabe, sharing knowledge of special places in the landscape likely occurred through one’s social and familial network. The site thus turned out to be an excellent opportunity to examine current reconstructions of Terminal Archaic and Early Woodland lifeways in the Great Lakes Region and what is known about human and animal cremation, caching behaviour, the manufacture and deposition of copper tools, and ceremonial sites in general.
The partial biface fragments (tip, medial, and basal sections) are believed to have been originally placed as complete bifaces in one of the features. The presence of 57 complete tip fragments suggests that at least 57 complete bifaces had been placed in the feature.
Over 98% of the stone tools were manufactured of Onondaga chert and much of the assemblage (66 %) was thermally fractured as evidenced by surface discolouration of the chert and by the high incidence of pot lid fractures. This resulted in fragmentation of the bifaces; 99% of the bifaces are incomplete because of shattering.
The fragmentary and burned nature of the bifaces made it difficult to date the site or identify the Indigenous people that lived there. The shape and size of the bifaces are consistent with tools found in Early Woodland Meadowood caches around 3000 to 2000 years ago, but Terminal Archaic biface preforms, from the era immediately preceding it dating to approximately 4000 to 3000 years ago, are similar in shape and size. The bifaces are unlikely to have been manufactured on this site since only 428 pieces of debitage were recovered from surface collection, block excavation, and feature excavation. Debitage are the waste flakes left behind when stone tools are created, and we would expect to see significantly more if the bifaces were being manufactured or refined on site.
Copper beads, one complete copper awl/pin, one copper awl/pin tip and one copper fishhook were also recovered (Figures 3 and 4). Although some of the beads are fragmentary and corroded, the vast majority are complete. They are exceptionally similar to those found at the large Middlesex Boucher cemetery in northern Vermont and at the Glacial Kame Hind site in southwestern Ontario. They were created by rolling a hammered strip of copper and overlapping one end over the other.
Trace element analysis has been conducted on copper tools and pieces from several Late Archaic and Early Woodland sites indicating that the region surrounding Lake Superior, especially the Keweenaw Peninsula of Michigan and Isle Royale, where open pit mines have been documented, was the source for most of the copper for the Northeast during the Late Archaic and Early Woodland periods.
Given the absence of any scrap copper, the copper tools, like the bifaces, were likely manufactured elsewhere, either at the source or at a macroband settlement, and deposited at the site. As 21 of the beads were found in the feature with the bifaces, it is also likely that they were placed there and then dispersed later by ploughing.
Pins, needles, and awls, variously labelled based on size, have been found on “Old Copper Complex” sites in the northeast and at several Terminal (Small Point) Archaic Glacial Kame cemeteries usually as grave inclusions. The copper fishhook was manufactured from a single piece of native copper of approximately 4.5 mm in maximum diameter at the shank tapering towards the eye and the point. Copper fishhooks, like the Graham specimen, are found throughout the Great Lakes area and date from the Late Archaic period to the Middle Woodland period.
In an effort to more accurately determine the temporal and cultural affiliation of the site, AMS radiocarbon assays on three butternut fragments (Juglans cinera) and a small sample of wood charcoal were undertaken, yielding when calibrated, a tight cluster suggesting a single occupation between 3061 and 2982 BP (Before Present) at 68.3% probability and between 3155 and 2892 BP at 95.4% probability. This places the site in the transition from Terminal (Small Point) Archaic to Early Woodland, likely around 3000 years ago.
The Terminal (Small Point Archaic) in southern Ontario has been dated around 3500 to 2800 BP and Hind specifically at 2900-2700 BP. The Terminal Archaic Glacial Kame and Haldimand burial complexes have cemeteries that exhibit similar practices to those of the succeeding Early Woodland Meadowood and Middlesex complexes, indicating a continuum from Terminal Archaic Haldimand and Glacial Kame through to Meadowood. Graham thus represents a moment in the local southern Ontario transition between two broad eras: the Terminal Archaic (e.g., Glacial Kame) and Early Woodland Meadowood periods.
The regional population to which the Graham Site people belonged is not well known. While no Glacial Kame sites are known within the Credit River drainage, fourteen Early Woodland sites have been documented, most of which are campsites or isolated finds located in the lower reaches of the river.
The Graham site, therefore, seems to represent a Terminal Archaic-Early Woodland ceremonial site (ca 3000 BP) in the upper reaches of the Credit River, likely associated with a family that was part of a seasonally dispersed band. The cremated animal parts, some identified as deer, along with ritually “killed” bifaces and copper beads and tools, suggest that a feast was held at the death of a family member with the remains of the deceased carried to a final cemetery.
This interpretation of the site would underscore Carloyn King’s long-standing observation noted above that some small sites, in this case, an eight-artifact surface scatter and four underlying features, can be as significant as large ones.
To see a detailed analysis of this site and its material culture, see “Minidoo’win’keng: Archaeological Investigations of an Ancient Ceremony on the Credit River in Southern Ontario” in Ontario Archaeology, Volume 102, just released.
Written by Ronald F. Williamson
Founder and Senior Associate
Acknowledgements (for the published paper)
I would like to thank Carolyn King for a long friendship and for providing the correct Anishinaabemowin word for ceremony, and for reviewing the premise and content of this paper. I would also like to thank Suzanne Needs-Howarth and Susan Pfeiffer for their recent reviews of the cremated bone assemblage, Jessica Lytle for her work in identifying wood charcoal and butternut remains from macrobotanical and flotation samples from the site, Jennifer Birch for coordinating radiocarbon analysis and for modeling the dates, and Debbie Steiss, and Doug Todd for their reviews of the lithic assemblage. John Howarth photographed the bifaces and undertook the digital enhancement of other photographs. Andrew Stewart prepared Figure 1 and provided much appreciated comments on an earlier draft of the paper. Martin Cooper, Danielle Macdonald, and Bruce Welsh aided in the preparation of the original report writing and Caitlin Coleman (who also helped with simplified terminology for this blog) and Andrea Carnevale provided logistical support in locating the artifacts and original ASI analyses for the site. I am grateful for the very helpful comments provided by the two reviewers for the paper, Chris Ellis and William Fox
