The Davidson Site (AhHk-54)

Test excavations, Davidson Site, looking southeast, June 13/06. Red stakes represents 200E line of grid system. Wooded/brush area in centre background is location of 1978 salvage excavations at the site by Ian Kenyon (1979).


This webpage describes some of the results of my most my most recent fieldwork. As a university professor I work for the public good by contributing to human knowledge and specifically, as an archaeologist, to our permanent, well-documented, knowledge of the human past. Therefore, I feel it is my duty to publicize the work I carry out and I will continue to do so in public forms as via this website. Indeed, major granting agencies are actually now requiring one make available the results of research they fund through venues such as websites. I often do so, however, reluctantly because by publicizing my discoveries and those of my colleagues, some will continue to use such information to essentially destroy the archaeological record via means such as unauthorized relic collecting.


One of my long-term research goals has been to investigate Late Archaic (ca. 5000-3000 years ago) settlement systems, and examine their broader anthropological implications for documenting and understanding variability in such systems amongst hunting and gathering peoples. The Davidson site was investigated beginning in 2005 and work is still ongoing to address these issues. In particular, the work has attempted to address some nagging doubts I have had about our dominant characterization of the settlement and subsistence patterns of these Late Archaic peoples.

Work since the 1960s has tended to see Archaic pre-agricultural hunting and gathering peoples through a progressivist or evolutionary lens: the became more "complex" including a lessened annual settlement mobility (e.g. fewer residential/camp moves) through time from the Early to Middle to Late Archaic (e.g. Ellis et al. 1990). However, even Late Archaic peoples were seen to have a relatively high degree of settlement mobility and it was argued that real settlement stability/a greater degree of sedentism only came in with the Middle Woodland (ca. 2000 BP) where very large sites and substantial true midden deposits are documented (e.g., Ferris & Spence 1995:98-100; Lovis et al. 2001:628-629; Monaghan & Lovis 2005; Wilson 1990, 1991).

I believe much of our "understanding" of the Archaic is driven by theoretical precepts as the available data to actually rigorously evaluate ideas is, in many cases, sadly deficient or lacking (Ellis et al. 2009b:827-828; see also Emerson and McElrath 2009). One would be hard pressed, for example, to provide much good data that Late  Archaic peoples were less mobile than Middle Archaic ones. Indeed, if we are ever to truly document and understand patterns of Archaic settlement we need to find and investigate sites that 1) have good floral and faunal preservation that can inform us about season(s) of site use and subsistence practices and 2) are minimally disturbed by ploughing and have preserved cultural features like pits and houses. Moreover, newer theoretical perspectives that emphasize historical processes and human agency, have begun to replace progressivist evolutionary ones that had dominated the Archaic landscape over the last 30-40 years (e.g., Emerson and McElrath 2009; Sassaman 2010, 2011). These kinds of perspectives require we have accurate chronologies that allow us to measure time at very small scales. In short, we need to excavate sites that can provide multiple dates for occupations from good stratigraphic and feature contexts.

My initial work at Davidson strongly suggested that it had large areas of intact deposits below the ploughzone including large numbers of significant preserved features and, although preservation was not perfect, the deposits did contain preserved organic remains or copious amounts of datable materials. These characteristics suggested it could truly be used to test ideas, rather than speculate, about the Late Archaic and they justified the extensive work that has been carried out at the site since 2008.

The Davidson Site

The Davidson site is located on the Ausable River just inland from modern Lake Huron near Grand Bend, Ontario. It is situated just above where the river enters a large wetland/marshland that bordered the south shore of the lake.  Although it had been known to local relic collectors since at least the late 1940s, the Davidson site was first reported to a professional audience by the late Ian Kenyon (1978, 1979, 1980a, 1980b). Kenyon (1978) discovered a buried component at the site during a fall canoe survey along the Ausable in 1977. The river had eroded a large section along the bank that had cut through the wooded area lining the river itself and extended into an adjacent cultivated field (see map below). This event exposed archaeological materials, some of which were in a buried "black humic" layer of about 10-15 cm thick that represented the "A" horizon of an old buried soil or "paleosol."  The  paleosol was as deep as 1.5+ meters deep in areas closer to the river but away from the river and into the cultivated field (southeast along the eroded bank) it got shallower and shallower such that it ended up partially incorporated into the ploughed zone of the field and eventually merged completely with the ploughzone in some areas.


Davidson site map as of December 2013 showing surface finds and excavated areas. Eroded bank area that initially led to discovery of buried deposits by Ian Kenyon was located just to south of excavated area and is indicated by large red arrows. Figure by Jim Keron.

In examining the buried black humic layer a Late Archaic Broadpoint (see  examples of this point form pictured below) was recovered and several cultural features such as pits were discovered eroding out of the wall.  In 1977, Kenyon (1978) cleaned up the wall including the exposed features. One feature yielded charcoal and a date of 3780 +/- 85 B.P. (calibrated to 2400-2030 B.C.) was obtained, the first radiocarbon determination on an eastern Great Lakes Broad Point Archaic site. Additional salvage excavations were carried out along the eroded bank in 1978 (Kenyon 1979), work that uncovered several other features including at last one Middle Woodland one. Stone artifacts, including numerous Broad Points and unfinished bifaces ("preforms") on Onondaga and Kettle Point chert, as well as coarser-grained metasediment rocks, notably sub-greywacke, were reported. These were compared to the Genesee point style first defined in New York state. Fragmentary calcined and uncalcined animal bone was associated with the Broadpoint component, some of which could be identified as deer, softshell turtle and dog. Also recovered, was black walnut shell and wood charcoal amongst which oak was the dominant species represented.


Davidson stemmed points and preforms, 1970s work. 1-4: Onondaga chert, 5: subgreywacke, 6-8: siltstone. From Kenyon (1980a).

From the top down, Kenyon (1978, 1979) described the stratigraphy along the riverbank as follows (see profile diagram below). The uppermost layer was, depending upon horizontal location, the ploughzone or modern topsoil (Ap) which he called zone 3b or a Dark Silt . As noted above, this layer was immediately underlain by a remnant of the black humic layer or top of the paleosol (Ab) farther back from the river or farther east and south along the eroded bank. However, as one moves closer to the river the ploughzone/topsoil is underlain by what  Kenyon (1978) described as a yellow and grey silt or light silt (his zone 3a) and that we recognize as a B1 and B2 soil horizon. Kenyon (1978) attributed this zone to flooding during a slow water phase of the Ausable. He also believed the flooding events responsible for depositing these sediments represented historic events caused by extensive European land-clearing, erosion off of fields into the nearby river, as well as damning downstream. At least closer to the river, underlying Zone 3 was the paleosol itself that Kenyon (1978:3) called Zone 2 and believed was developed on sands representing deposits of the Nipissing phase, a high water level in the lake Huron basin that had flooded the site area prior to ca. 4500 B.P. However, our research suggests that these are actually deposits formed by overbank flooding of the river that could pre-date the Nipissing Phase. Kenyon divided this zone into three subdivisions representing the three horizons of that old soil. It included the 10-15 cm thick, black, humic "A" horizon (Zone 2c; our Ab), the underlying 20 cm zone of accumulation, or  our "Bb" horizon, that was described as an orange sand (Zone 2b), and the underlying unmodified "C1" horizon (Zone 2a) described by Kenyon as a "yellow sand".

In addition to the material by the eroded bank, which he called Area A, Kenyon (1979) also found cultural debris on the ploughed field inland to the north and east that he called Area B and noted that it included materials relating to later occupations at the site. However, he did not investigate this area in detail


Profiles of Davidson site areas near river bank. Excavation profile from 2008 work (left looking south as described by Andrew Stewart) is compared to Ian Kenyon's description of these layers from the 1970s (right). Note buried paleosol/black humic layer buried under sterile flood deposits (B1, B2, 3a) extending down from ploughed zone (Ap/3b)


2005-2014 Investigations

I began examining the site in 2005. Initial surface examination of the ploughed field revealed Broadpoint related material over a wide area inland or back from the river. Especially common were numerous large thinning flakes on the coarse-grained material called sub-greywacke, derived in making the large Broad Point bifaces. As Kenyon (1980b) argued, in areas such as the Ausable there are no sources of large relatively flaw-free pieces of cherts needed to efficiently make the points. The local major source, Kettle Point, consists of cherts that tend to be smaller in size and relatively flawed. As a  result, Broadpoints in the area were often made on cherts imported from other areas that do occur in larger and flaw free packages, such as Onondaga from nearer Lake Erie, or on other coarser-grained rocks such as sub-greywacke and siltstone that, although harder to flake, occurred as large flaw-free cobbles in secondary sources such as local river beds and glacial age deposits. Therefore, the large sub-greywacke and other flakes on coarser-grained rocks from making such tools are just as diagnostic of Broad Point components as the points themselves. Broadpoints on Kettle Point chert do occur. In fact, Kettle Point chert flaking debris, including numerous large thinning flakes, dominates on such sites in the area. However, Kenyon (1980a, 1980b) argued that points on that chert are often smaller and more atypical in outline shape because of the limitations imposed by that raw material. Significantly, little artifactual evidence was found on the ploughed surface in 2005 adjacent to the river immediately north of the eroded area where Kenyon (1978, 1979) had worked and found the buried and sealed level, suggesting the old paleosol was present but still buried and intact in that area.

Since 2006 we have carried out a program of test pitting/smaller scale excavations (2006, 2010, 2014), coring (2007, 2010, 2014) gradiometer/magnetometer surveying (2008-2013), magentic susceptibility surveying (2013), surface mapping (15 complete surface mappiings from 2007 to 2014) and, during the 2008-2009 summer field seasons, 31 weeks of larger scale excavations (see figures below; Eastaugh et al. 2013; Ellis 2006, 2010, 2011, 2013, 2014; Ellis and Eastaugh 2014; Ellis et al. 2009, 2010, in press). Specialists also are involved in ancillary geological studies. This work, supported by two grants from the Social Sciences and Humanities Research Council of Canada (SSHRCC; Grant #410-2007-1690; Grant #435-2013-0324), has revealed that the site covers ca. 1.9 ha, placing it spatially amongst the largest Archaic sites ever reported for Ontario. The old site surface is deeply buried over a large area, especially closer to the river. However, our work has revealed that the old paleosol is at least partially intact even well out into the ploughed field and has to cover at least 3500 m2.  While distinctive  Broadpoint Archaic (4000-3400 BP by radiocarbon; calibrates to ca. 2500-1500 BC in our calendar) artifacts are plentiful and widespread over the whole site, our work has also demonstrated that there is a significant, subsequent, and, while more spatially restricted to the north end of the site (see distribution of red stars on map above), a quite intensive occupation during the Terminal or Smallpoint Archaic (ca. 3400-2800 BP; calibrated ca. 1500-800 BC).  Several surface finds of these small, usually notched or expanding stemmed, points were recovered in surface collection and excavations (see pictures below). Post-2800 BP materials of the Early (Meadowood), Middle and Late Woodland are present as well but these are rare and indicate only intermittent and ephemeral site use during those later time frames. 


Davidson site map showing excavated areas (in red), and areas surveyed with a gradiometer/magnetometer as of 2013. Note dense band of positive magnetic anomalies running diagonally from southwest to northeast across northern end of surveyed area. Map by Jim Keron and Ed Eastaugh.

Also of signficance in delineating the site is the patterning revealed by the gradiometer surveys carried out from 2008 to 2013 (see maps above and below for survey locations). Such work was carried out under the direction of Edward Eastaugh of the Department of Anthropology at Western. The equipment employed was the UWO Anthropology department’s Geoscan FM256 Fluxgate Gradiometer (2007-2011) and Sustainable Archaeology's Bartington Grad 601 dual sensor system (2012-2013). This equipment measures minute differences (anomalies) in the earth’s magnetic field caused by buried archaeological features such as refuse pits or piles of fire-cracked rocks (FCR). The surveys were conducted within 20 x 20 m (400 m2) grid blocks and readings were logged at 0.125 m intervals along parallel traverses spaced 0.25 m apart. As shown on the accompanying maps, these combined surveys reveal a dense band of positive magnetic anomalies about 25 m wide and 160 m long extending away to the northeast or diagonally from the current river course. The 2013 magnetic susceptibility survey reveals a similar pattern, albeit at a grosser scale with a dense area of higher magnetic susceptibilty in the same area as the gradiometer anomaly band (see map below). 


Davidson site gradiometer survey as of 2012 (left) showing band of positive (black) magnetic anomalies probably paralleling old water course of river. At right is magenetic susceptibility mapping completed in the spring of 2013 that shows very high readings in the same area as the dense band of gradiometer anomalies. The areas encompassed in yellow at right seem to correspond to sediments left by overbank flooding since European land alterations in the 1800s that have buried/mantled the old land surface along the river containing the archaeological deposits in that area. Maps by Ed Eastaugh.

Ground truthing through excavation of several of the larger gradiometer anomalies has revealed they are all significant cultural features such as houses and large pits or pit clusters -- there must be literally hundreds of cultural features at the site (Eastaugh et al. 2013). The dense diagonal and linear concentration of these features corresponds to the area of surface finds (there are almost no surface finds due north of where we excavated in the area between the "anomaly band" and the current river) and strongly suggests the inhabitants were camping adjacent to and lining the river at the time the site was occupied. Therefore, the river was located east of its present location in Late Archaic times and has since moved west until becoming entrenched in its current position. Preliminary results from ground penetrating radar work by Roger Phillips of the Department of Geography at the University of Toronto reinforce this idea as the deposits between the current river and "anomaly band" were deposited differently than in the band location itself.


Davidson stemmed Broadpoint bifaces, ca. 3800-3400 BP, Onondaga Chert.

Davidson Broadpoint bifaces, ca. 3800-3400 BP, Sub-greywacke and other coarser-grained rocks. A-C: Points; D-F: Preforms (bifaces discarded in manufacture); G-H: Bifaces with round and polished tips, possibly from hide-working.

Davidson Terminal/Smallpoint Archaic Stone Points, ca. 3400-2800 BP. A-D: Hind style points; E: Innes style point; F-H: Crawford Knoll/Preston-Notched-like points.

As the anomalies in the main excavated buried area near the river proved to be cultural features we reasoned similar, but rarer and usually smaller, anomalies located away from the band lining the old river course were also such features and notably in the southern half of the site where only Broadpoint Archaic materials had been found. To test this idea in 2014 we excavated two small areas in the ploughed field in locations where there are no overlying flood sediments, targeting anomalies in those two areas that corresonded to surface concentrations of Broadpoint bifaces and coarse-grained flakes thinning flakes. In all cases we found plough-truncated Broadpoint feature remnants corresponding to the anomalies in those targeted units (see photo below).

Bottoms of two plough-trunated, overlapping, Broadpoint, pit features uncovered in 2014 that corresponded to two adjacent gradiometer anomalies in the southern area of the site.

Our main site excavations in 2008-2010 were concentrated in the area just north of where Kenyon (1978, 1980a) had worked.  As a whole the excavation grid was oriented with Grid North at 30o west of magnetic north to conform to the orientation of our original 2006 test units off the ploughed agricultural field and it is grid north that is indicated by the arrow in th accompanyting photographs. Several separate excavation units, referred to by various terms such as Block Unit A, Anomaly Area I, etc. (see map below) were opened up. These units were selected as potentially very productive areas based on the coring, test unit excavations and gradiometer surveying. For example, Block Unit A was placed in an area that the gradiometer survey revealed had significant anomalies and that probing showed had thick organic deposits/large features while Block Unit B was excavated because a one by two metre test unit opened up in 2006 revealed very thick deposits including at least two large pit features with Broadpoint diagnostics. Excavations actually proceeded in 50 cm2 quadrants (see picture below). One metre units were referred to by the the intersection co-ordinates of the grid lines at their southwest corner and the 50 cm quadrants in each were referred to by their relative position within each one metre unit (e.g. SW, SE, NW, NE quads).



Areas excavated 2006-2010. Main grid is in five metre blocks.


Excavating 50 cm quads in arbitrary contoured 5 cm levels into buried paleosol A horizon (Ab), Block Unit A, 2009, looking west, Davidson site.

Excavations in 2006-2010 totaled only 84 m2. So only a fraction of this huge site was excavated over the 35 weeks devoted to that activity. The excavations were limited because of the complexity of the deposits and the large numbers of artifacts and features we uncovered that needed to be carefully excavated and recorded. Although dominated by non-descript flaking debris and fire-cracked rock (FCR) fragments, tens of thousands of artifacts have been recovered that fill over 85 bankers boxes, an unprecedented amount of material for limited excavations on an Ontario Archaic site. Although bone and antler preservation is not perfect, faunal remains such as deer, bear, beaver, canid, turtle shell and even fish remains were recovered as was the odd organic artifact such as a harpoon tip and a bone point. In addition, there is good preservation of floral materials such as nutshell, seeds and charcoal. Stone tools beyond points and preforms are surprisingly rare. We did find at least one copper artifact, probably a small fish gorge. The artifactual material is still being processed and has not been subjected as yet to very detailed analyses. In the small area excavated, which corresponds to the south end of the band of anomalies revealed by the gradiometer survey, we uncovered no less than 85 culturally significant features including some very large ones, notably houses. There must be thousands of features at the site.

2008 Excavations in Block Unit "B", Davidson Site. Workers, including University of Western Ontario graduate and undergraduate students, are exposing the top of the buried black layer.


Sub-greywacke Genesee Point in situ in "black humic layer" (Ab), Block Unit B, Davidson Site. The arrow used in all photos is 20 cm long and is always oriented for convenience facing grid north.



As noted above, we have discovered a large number of significant cultural features at the site, many of which are feature types never or arely reported at this early date. As the site was intensively used over a thousand year period at the end of the Archaic, or during the Broadpoint and Smallpoint Archaic, it is not always possible to assign features to specific time periods. Indeed, one could have a later Archaic group use feature depressions left by earlier occupants to discard refuse and digging of later pits could also incorporate older artifacts. Nonetheless, careful consideration of feature contents, horizontal spatial custering and stratigraphic relationships of these features, along with the dating by radiocarbon (AMS) of many of them (25 AMS dates now available), has allowed us to assign most examples to use during the Broad Point Archaic (3800-3400 RCYBP;ca. 2500-1700 cal. BC), Terminal/Small Point Archaic (3400-2800 RCYBP; ca. 1300-1100 cal. BC) and in rare cases, probably the Early Woodland (ca. 2800-2400 BP; < ca. 1100 cal BC).  Plotting of the dates, including the one date obtained by Ian Kenyon in the 1970s, as summed probability distributions actually suggest three major periods of site use (see diagram immediately below), the two earliest of which (ca. 2500-2300 BC and 1800-1700 BC) are Broad Point and the third of which is Small Point. Some of most significant feature of the Archaic period are described below.

Summed probability distributions of Davidson site radiocarbon dates. Red area indicates one sigma distribution and the red plus blue indicates the two sigma distribution.

I note that most of the dated features and feature clusters, and especially the large ones that showed up clearly in the gradiometer survey seem to date to Terminal Archaic Small Point rather than Broad Point times. This assignment also agrees with the surface distributions of diagnostics. As noted earlier, reference to the plotted surface finds mapped above shows that Broad Point diagnostics, such as the bifaces and sub-greywacke/coarse-grained flakes are distributed over a wide area that goes much beyond the the band of large anomalies lining the former river and even extend far to the south and back from the river location at the time. In contrast, the Smallpoints are concentrated more in the anomaly band or where the large features are concentrated, nearer the river.

Broad Point Archaic Features

At least 14 definitive Broad Point Archaic features were found in the 2008-2010 excavation, most of which are smaller pits probably used for storage or caching -- although like most features, these surface depressions often ended up being used simply as convenient places to dispose of garbage once they had outlived their usefulness in their primary function(s). Nonetheless, some significant Broadpoint features were uncovered although unlike the Terminal Archaic occupation, no definitive houses or other structures clearly assignable to that occupation have been discovered.

One notable Broad Point Archaic feature was a large storage pit (Feature #1), about 1.5 m across and almost a metre deep excavated in 2006 and 2008 in Block Unit B. AMS dates of 4020 +/- 30 BP (Beta -381334) and 3660+/-30 BP (Beta - 294186) on wood charcoal from the bottom of that storage pit is certainly consistent with a Broad Point affiliation (the diference in dates could be due to dating old wood), as is the fact it contained several actual Broadpoints, including one worked into a drill, and several sub-greywacke flakes. A large number of very large biface thinning flakes on Kettle Point chert,  undoubtedly from Broad Point preform reduction, were also included. Less diagnostic tools include a small stone celt or ungrooved axe and the end of what is apparently a bone bipointed gorge used in fishing. The faunal and floral material from the pit have not been fully analyzed but deer, turtle shell and fish bone, including a definitive ear bone or otolith from freshwater drum/sheepshead, were in the fill. Floral materials include charred butternut and black walnut shell as well as seeds of raspberry, grape, unspecified cherry as well as unspecified prunus (plums or cherries) and cleavers. Identified wood charcoal includes maple, beech, ash, elm and red oak, as well as some unidentifiable conifer.

Large storage pit (Feature #1) filled with refuse, 2008 excavations, Block Unit B. Several Broadpoints, including one remade into a stemmed drill, were recovered from this feature as was a small, ground stone, celt/axe.

Another Broad Point feature of note was discovered in the west end of what we had called the east-west trench (see map above for location). This feature was a large natural erosional channel (up to 80 cm deep and almost 2 m across), with one or more smaller feeder channels (see plan map below). We have excavated a three metre long segment of this channel but it undoubtedly extends some metres farther to the south. The channel had been almost completely filled with cultural debris and clearly represents a midden or specialized garbage disposal area. In some areas, several distinct layers of cultural infilling are represented suggesting it was used as a dump for more than one period of time although radiocarbon dates (see below) suggest use of this midden was within a fairly restricted period.

A small pit (Feature #17A) had been cut down into the top of the Feature #17 midden deposit and it contained a Terminal Archaic Crawford Knoll point that should date to ca. 3400-3000 BP. Another Terminal Archaic point was found laying right on the top surface of the midden. However, the rest of the deposit yielded only Broadpoint diagnostics including several bifaces and many subgreywacke flakes.  Another pit containing Broadpoint diagnostics (Feature #18) was uncovered that also cut through the edge of the underlying Feature #17 midden and that, in turn, had  been cut through by the Terminal Archaic Feature #17A pit (see illustration and photo below). Charcoal from Feature #18 was dated to 3400 +/- 40 BP (Beta – 257095) indicating the underlying Feature #17 has to be at least that old. Charcoal from near the top and bottom of Feature #17 itself also was submitted for radiocarbon dating. This charcoal from 15-20 cm deep in the midden in unit 198N/207E was recovered from the same level as a Broadpoint stemmed drill. It was dated to 3450 +/- 40 B.P. (Beta - 277023). Charcoal from nearer the bottom of the same midden at 60-65 cm deep in 199N/207E that was recovered right beside the tip of a large sub-greywacke biface produced a statistically identical date of 3480 +/- 30 B.P. (Beta - 381335) confirming that the midden as a whole was in use for a short period of time until just before 3400 B.P. These dates all fall within the second period of Broad Point use of the excavated area suggested by the overall radiocarbon date distribution pictured above.



Plan map of East-West trench (north of 200N) and Feature #17 Extension Area (198N to 200N) excavations. Extent of feature #17, a natural erosional channel filled with refuse, is shown in blue. The feature undoubtedly extends for several metres to the south and is a true Broadpoint  garbage dump or midden used for the organized disposal of refuse.


Refuse-filled natural depression or midden (Feature #17) in east-west trench along 200N line looking south, 2008 Excavations. Large chopper lying on bottom of depression can be seen protruding from wall at left. Broad Point pit feature (Feature #18) is in foreground right. The depression, and the adjacent pit feature, contained definitive Broad Point bifaces. Feature 18 is dated to 3400 +/- 40 BP (Beta – 257095) and two separate charcoal samples from Feature #17 to the south were dated at 3450 +/- 40 B.P. (Beta - 277023) and 3480 +/- 30 B.P. (Beta - 381335) .

A  final notable feature form associated with the Broad Point component is pit hearths. A small example containing fire-reddened soil and a fire-damaged Broadpoint base/stem was located in the 2010 work along the river (see picture below) where it had been partially truncated by a Terminal Archaic pit and a pithouse that are directly dated to ca. 3000 B.P. A second pit hearth that was somewhat larger and had a considerably fire-hardened and reddened lining was found in 2014 at the very southern edge of the site where we targeted magnetic anomalies via limited excavations.

Pit hearth (Feature #73) with fire-reddened soil containing Broadpoint stem as first exposed in excavation, 2010.

Terminal Archaic Features

A  number of significant features have been uncovered that date to what is traditionally regarded as the Terminal or Small Point Archaic of ca. 3400-2800 BP. As is the case with Broad Point finds, many of these are typical smaller to medium sized pits that often occur in clusters. While there may be contrasts between clusters, within each cluster the constituent pits are often of the same size and morphology and one presumes, function. This spatial correlation suggests the pits in a cluster represent an “integration of events” (Holdaway and Wandsnider 2006; Wandsnider 2008:62). In sum, and as argued by Bettarel and Smith (1973 15-19) and Binford et al. (1970:61-66) each integrated cluster represents a short period of time in which a specific area was repeatedly used for the same purpose by people who probably knew each other on a face to face basis. Radiocarbon dating of multiple features within particular Davidson clusters has tended to confirm use of all features in a cluster about the same time. Aside from the pit clusters, there are some other very notable types of features, namely structures including houses. A forthcoming paper provides additonal doumentation of the probable houses (Ellis et al.. 2015), but the documented examples are summarized below beginning with those that date most recently in time.


Feature #3/9

One structure is a small house pit with sloping entrance on its east end (Feature #3/9) that was initially investigated because it registered as a significant positive anomaly in the gradiometer survey (see photos and drawings below). Located in what we called Anomaly Area I (see map above for area location), this house was at least 4 metres by 3 metres in size. Based on outlying posts, only the central ca. 2 by 2 metres metre area seems to have been shallowly dug down into the ground with the house edge a flat shelf at closer to ground level. Suggestions of similar structural forms (e.g. with only the centre of the house dug down into the ground), albeit dating much later, have been reported elsewhere (e.g. Binford et al. 1970: Figure 8). The interior and exterior posts were quite substantial (12 to 15 cm across). In the center of the house was a central pit with some evidence of fire-reddening suggesting it was a hearth. Charred black walnut shell from the floor of the structure adjacent to the central pit yielded a date of 2850 +/- 40 BP (Beta-257094).


Plan map of Feature #3/9 house showing central pit (#3A), confirmed and possible post locations and later dating pit feature overyling the house (Feature #11). The outer posts indicate the house floor was bigger than simply the central area (Feature #3/9) that had been dug down into the ground. Those exterior posts are substantial and may represent support posts rather than wall posts -- if so, the house could have been bigger than the area excavated. Charred black walnut shell from the floor right beside the south edge of the central pit (3A) was dated at 2850+/-40 BP (Beta-257094).

Terminal Archaic pithouse (Feature #3/9) in "Anomaly Area I", 2008 excavations, looking east. Round rock at front left is in probable post stain

A large mounded area of black soil overlay this house and presumably reflects a collapsed sod/soil roof that are typically used historically to cover and insulate such structures (e.g. Ames and Mashner 1999:151-152). In fact, although not always easy to trace, the deposit seemed to contain a darker upper blacker "core" of deposits (see profile views below) that could be distinguished from the lower lighter coloured deposits and which was actually separated in some areas, such as near the bottom of the central pit and just inside the house entrance, by a thin area of sterile yellow soil. One suspects the sterile yellow intrusions represent sands washed onto the floor area in the short time between abandonment of the house and the roof collapse. The upper core contained many more artifacts than the lower fill of the house itself and even included sub-greywacke flakes. Presumably these earlier artifacts are accidental inclusions taken up with the roof sod/soil collected from surrounding areas.


Profile of  pit house (Feature #3) in "Anomaly Area I" along 209N grid line. Note the central pit, roof support post #12 and sloping entrance at right, 2008 excavations. Yellow lines demarcate darker "core" of deposit that probably represents collapsed sod/soil roof  deposits.

View in partial plan and profiles of northwest quadrant of central pit (Feature #3A) that underlay pithouse Feature #3/9. Note thin layers of lighter, sterile, yellow soil in middle of feature and adjacent pithouse itself (indicated by yellow arrows) that separates "floor" deposits from darker, overlying, collapsed roof deposits.

Close up of some posts (#12 and #6) associated with Late Terminal Archaic house (Feature #3/9) in Anomaly Area I, 2008 excavations.

Plan view of central support post (#11) underlying southern edge of lip of central pit in Feature #3/9 pithouse. Note thin yellow layer of sterile soil in wall at left of post separating upper darker "roof" deposits from lower  slightly lighter coloured "floor" deposits.


Feature #32

The entrance end to another house (called Feature #43), that is similar to Feature #3/9 and presumably of a similar age, was also encountered just to the southwest in the east edge of what we called Block Unit C. In that block the house cut through, and thus must post-date, yet another house called Feature #32 (see plan drawing below). Feature #32 was ca. 3 metres  in diameter and was detectable because it had been outlined by a 25-30 cm wide and about 25 cm deep wall trench excavated to insert the wall posts. As the posts themselves are inserted largely into the dug trench and did not extend much below its depth, they were difficult to detect and often visible only as slight "bumps" extending just below the trench. Nonetheless, the very bottom of some of these posts were documented as shown on anaccompanying illustration below. There was a gap in the west edge of the wall trench "ring" feature facing the river that clearly represents an entrance and two large roof support posts (Posts #23 and 24), 12 to 13 cm across and 35 to 40 cm deep, were in the centre. Although the floor of the house was intact, there was no evidence of any internal hearth and the only possibly associated features were a small pit (Feature #63) located by one of the interior support posts and a shallow larger pit near the rear (east side) of the structure (Feature #61).

As the south and north edges of this house cut down into deposits that are radiocarbon dated to ca. 3200 to 3000 BP (see below), the house could date to as early as ca. 3000 BP. Charred black walnut shell from the bottom of the northwest segment of the wall trench was submitted for dating and the result of 3050+/-50 (Beta-294187) is consistent with that assessment. Interestingly a Late Archaic house outlined by a wall trench had been reported earlier from the Rock Hearth site in southern Michigan (Garland et al. 1990:176-179). While its interior and part of the wall trench were partially destroyed by recent activities and some flooding, that house seems to have been of a similar size to the Davidson example. The Rock Hearth house may have had a more squarish overall shape with rounded corners but like Davidson it seems to have a gap or entrance on one side,  a wider and abrupt narrowing of the trench to the right of the entrance when viewed from the house interior (seen just northeast of Post #21 on the map below), and shallow and difficult to detect posts. The Rock Hearth structure also seems to date earlier than Davidson, having two radiocarbon dates of around 3700-3800 BP. Nonetheless, it suggests houses of this nature may have been relatively widespread and used for a fairly considerable period of time during the Late Archaic.


Plan map of Feature #32 house outline by wall trench in upper part of Block Unit C. Note Feature #43 at right centre that cuts through and post-dates the wall trench house. Feature #43 represents the entrance to a pit house comparable to one almost completely uncovered in Anomaly Area I at the site (see description above).


Photo of gap/entrance to Feature #32 wall trench house.



Profile of Feature #32 house wall trench showing tips of Posts #41 and 42 protruding below bottom of the trench. Note underlying lighter grey Feature #65 pithouse deposits (see below) and their associated roof support Post #38. 


Feature #5/26/33A/65

Partially underlying the Feature #32 wall trench house is a large pithouse dug deeply down down into the underlying deposits.  This feature, called Feature #5/26/33A/65 (herafter Feature #5), was centered in Block Unit C but extended into the adjacent edges of Block Units A, B and what we have called the Block Unit C extension (see area excavation map above). As can be seen from an illustration below, this house was roughly circular in outline and about 5 metres in diameter. It extended about one metre deep and had a sloping entrance facing the river on the west side. After the house was abandoned it had become filled up by a combination of natural and cultural processes. Around 3000 BP, the sloping entrance had channeled rising river waters into the house interior on at least seven occasions and this flooding was interspersed with the deposition of darker black organic deposits containing some cultural debris. The result was a series of six thin organic layers separating the various flood events as can be seen in a profile drawing below. Note also the segments of the wall trench profiles of house Feature #32 in the overlying deposits.


Plan Map of Feature #5/26/33A/65 Pit House (yellow) in Block Unit C (200-204N/200-204E) and Adjacent Areas of Block A (area south of 200N), Block B (area north of 204N) and units excavated in 2010 (area west of 200E). Note eroded centre of house floor (light grey), bench around central floor area (blue) and remnants of interior linear partitions (green). Diagonal roof support posts are shown in red. Note other pit features in dark grey surrounding house, that date earlier (Features #73, 74, 76, 77) or later (Feature #33B, 35) than the pit house itself. Feature #72 at lower left was dated on ash charcoal at 2930 +/- 40 BP [Beta-294191] so is also Terminal Archaic in age.



North (left) to south (right) profile of Small Point Archaic pit house along 202.5 East line used between ca. 3200 and 3000 BP. Thin layers above the fill suggest that infilling of the house was interspersed with at least seven flood events (in yellow separating the thin grey layers above the bottom of the house).

Unfortunately, and as outlined in light grey in the plan map above, the flooding also eroded the centre of the house floor such that except for the very bottom of one central support post (Post #33), all central floor features/attributes had been removed. A thin layer lining parts of the house floor, especially in the uneroded upper entrance area, seems to represent a distinct actual compact floor or living surface. However, this was not visible to the naked eye in other preserved floor areas although to the touch the bottom of the floor deposits seemed more compact and "clayey" perhaps beause they contained more ash. Regardless, there was distinctive layers of debris in some locations that seem to relate to the actual use of the house. Notable here was a small deposit we called Feature #67 near the southeast corner of the house adjacent to the bench (see profile drawing above). That deposit could very well be refuse that accidentally collected in that area as such refuse does tend to collect along the edge of house floors.

The structure had around its north, east and south ends a 50 cm wide or so bench or shelf cut into the wall about 20-25 cm above the floor. This bench was built such that it terminated before reaching the west end of the block in the area where the entrance was located. There is some suggestion, notably on the south side nearer the top edge, of a more level surface or potential shelf in that area that could have been used for storage (see especially the southern part of the N-S house profile reproduced above @ 200N-200.5N).

Within the house there were several associated structural features. One obvious one is support posts for the roof. Six large (13-18 cm diameter) vertically oriented posts (see photo below) were found either on the floor or in the area of the bench/lower slope (see photo above where one diagonal post is visible just below the Feature #2 wall trench profile). More posts were probably  present in the front of the house where the floor was eroded. Diagonally oriented posts near the top of the slope that formed part of the actual sloping roof supports were certainly present. Three diagonal posts, #38 in the NE edge of the pit (see photo above of this post as it was visible under the Feature #32 deposits), #40 at the southeast edge and #51 at the northwest corner were documented but others may have been present in the area excavated, notably in the southwest area. However, the digging of later features could have removed them. Also, it is possible we missed some of these posts in excavating the southwest edge of the house. They tend to be shallower as they are not supporting the full weight of the roof and given their diagonal, rather than straight up and down, orientation they may have been dismissed as root or rodent intrusions prior to our recognizing this feature was a house.


Profile of large support post at back of bench in north edge of Archaic pithouse.

There was also clear evidence of partitions within the house in the form of shallow linear stains cut down below the floor and which it is suspected were used to insert something like bark flaps (in green on plan map above). One of these features (#68), was found running north to south along about 201.3E (see picture below). It began and was well-defined just inside the bench at about 203.5N, and extended south to about 202.6N, beyond which it became more attenuated and then disappeared due to removal by the flooding. This partition undoubtedly separated the front entrance half of the house from the rear. The other partition (Feature #70), about 8-10 cm wide, also ran north to south across the very back of the house at about the 202.7E line. It only hived off a small area at the back of the house, perhaps a small storage area.

Linear north to south partition remnant (Feature #68) across centre of pithouse. Note how it originates from beneath the floor of the house in the profile at upper right.

As for the age and cultural affiliation of the house, it is under the wall trench house dated at ca. 3000 BP so has to be older than that date. The question is, how much older is it and does it represent  Broadpoint or earlier Terminal Archaic use?  Decidous wood charcoal from within the hole for Post #32 at the back of the house, yielded a date of 3870 +/- 40 (Beta- 277027) initially suggesting a Broad Point Archaic affiliation.  A stemmed point from the bench at the north side would be consistent with that affiliation. Yet, there there is no evidence that the dated charcoal is from the post itself and indeed, there is no evidence the house burned down so it may be intrusive. Moroever, two other dates were also obtained on material from the thicker complex organic layers or sets of layers underlying the lowermost flood deposit. One of these dates was obtained on decidous wood charcoal from the deepest deposits just above the floor of the house (130-135 cm deep) on its south side, that we called Feature #67. As noted above, we believe the Feature #67 deposits is refuse collected at the edge of the floor along the edge of the bench. That date obtained was 3120 +/- 40 BP (Beta-277028) suggesting this is a Terminal Archaic house. Another date was obtained on charred black walnut shell from the north slope at 105-110 cm deep and produced a date of 3010 +/- 40 BP (Beta-277026), which also suggests a Terminal Archaic age. This date is actually statistically identical to the date on the overlying wall trench house above the flood deposits, suggesting those flood deposits accumulated in a relatively short period of time. At least two bifaces from within the house also appear to be roughly tear-shaped preforms, which are more consistent with a Terminal Archaic age and these later dates.

Small tear-dropped shaped biface in situ at back of bench at northeast side of pit house.  

Overall, we believe the house is of Terminal Archaic age and that the charcoal was accidentally incorporated in the post mould. The house had clearly been demonstrably dug through older Broad Point features including a refuse-filled depression/midden at the southwest corner and smaller features such as a pit hearth on the west side. Moreover, in excavating sizeable pit houses it is typical to throw up a circular ridge of earth around the house composed of the fill excavated from the hole and once the wooden roof is installed that earth, which could easily incorporate older materials, is stacked around and on top of the edges for insulation (see especially Hayden 1997: Figure 3.8). One expects the thick rim area deposits in the Davidson example (see especially the south end in profile depicted above) are the result of such practices. After the roof eventually collapses the older materials in the deposits can end up falling into the house itself. Regardless of the source of that dated deposit, the two dates on the thick deposits underlying the flood deposits suggest they are only about 3200-3000 years old. If the house was a Broadpoint one occupied at about 3800 BP and given the dates of  those 3200-3000 year old deposits on the fill right on the floor, one would have to assume nothing accumulated in the house depression by natural or cultural means for over 600 years, which seems highly unlikely. These two later dates therefore, most strongly suggest the house is of an early Terminal Archaic age at ca. 3200-3000 BP (Ellis et al. 2010).

Regardless of its exact age, this is a definitive Late Archaic house. Along with the earlier described Feature #3/9 and two small overlapping pit houses of a somewhat different style from the Thistle Hill site near Brantford apparently dated to about 3400 BP (Woodley 1990), they suggest this house form was perhaps commonly used in the Terminal Archaic of Ontario.

Other Structures

A final later Terminal Archaic feature of note is also a structure of some kind but it does not appear to be a house structure (Feature #22/64). This feature was uncovered in the east side of block unit A and extended east into what we called the Block A Extension. This feature consisted of a roughly oval platform (Feature #22A) about 2 by 2.5m in size outlined by several posts. The platform had been cut down about 20-25 cm into the underlying subsoil. Within the platform had been dug and redug several larger oval pits (see profile photo below). Only the very bottom was intact of some of the earliest deeper pits excavated in this cluster (e.g., Features #22C, 22E, 22F, 64). These pits had fairly uniform fills with the exception of the southwest edge of Feature #22B, which still retained an intact lining of FCR and which leads us to suspect these features were earth ovens. Feature #22 has been dated based on probable ash charcoal at 2800 +/- 40 BP (Beta - 277025).


Plan map of Block Unit A and adjacent Block Unit A Extension (area east of 204E line). Apparent Broadpoint Archaic features are in red while Terminal Archaic features are in green. The oldest cluster of Broad Point pits (Features #49, 50, 54, 31, etc. ) has so far yielded four consistent AMS dates on wood charcoal suggesting these features were used about the same time: Feature #49 has three dates of 3810 +/- 40 BP (Beta-294190), 3920 +/- 30 BP (Beta-0381342) and 3930 +/- 40 BP (ICA-C/0516) while the underlying Feature #50 is dated on wood charcoal to 3940 +/- 30 BP (Beta-381343). Other clusters of pits (see map above) do have Terminal Archaic dates of around 3000 BP [e.g. Feature #39: 2980 +/- 40 BP (Beta-294189 on Juglans spp. nutshell); Feature #29: 2930 +/- 30 BP (Beta-381337 on wood charcoal)] but they have also yielded statistically different dates consistent with a Broad Point age on charred nutshell [e.g. Feature #29 (3870 +/- 30 BP; Beta-381336) and Feature #37 (3830 +/- 30 BP; Beta-381339)]. These clearly inconstent dates that fit into two well-defined time periods strongly suggest the features in this cluster are Terminal Archaic ones that were dug into/through older Broad Point deposits and so contain a mixture of materials of two different ages.


Profile of Feature #22/64 structure along 204E line or east wall of Block Unit A. Note platform (#22A) into which large pits (#22B, 22D) had been emplaced. Feature #22 has been dated on probable ash charcoal at 2800 +/- 40 BP.

As these pits fill the structure it cannot be a house but must have served as a locus for a specialized cooking and/or processing function. It is notable that somewhat comparable large feature clusters, with platforms of debris into which pits and posts had been emplaced, were reported from the Late Archaic to Early Woodland components at the Peace Bridge site near Fort Erie (MacDonald and Williamson 1997:237-246). These features were seen as possible "living floors" or structures/houses and had a mix of Late Archaic point types. While comparable to the Davidson example they also seem to have a wider range of pit forms as well as hearths associated. Similar large, complex features, up to 2.1 by 3.3 m in size, were also reported  from the Feeheley site in Michigan (Robertson 1987:115). These are reported to be large organic stains within which one or more smaller pits were emplaced and which may also have associated post molds. The Feeheley site occupation is estimated to have occurred at about the same time as Davidson or at ca. 3250 to 2750 BP.

Feature #22/64 was surrounded by finds of Broadpoints and Smallpoints and some Broadpoints were actually recovered from its fill. However, these could easily be accidental inclusions as it had been dug down into older features of definitive Broad Point age such as a feeder extension of the earlier Feature #17 erosional channel filled with refuse/midden deposits dated to ca. 3400 BP+ (see plan map above). Also, it cut into an earlier cluster of overlapping features (Feature #28, 49, 50, etc.) that had Broadpoint diagnostics and Feature #49 has several dates at aro9und 3800 BP). Regardless, Feature #22 itself yielded a date on charred wood (probably ash) of 2800 +/- 40 BP (Beta-277025) consistent with its stratigraphic placement and a Terminal Archaic assignment.

As can be seen in the plan map above, there are several other Terminal Archaic features or feature clusters in Block Unit A. One notable Feature is #36. This pit, dated at 3020 +/- 40 BP (Beta-294188), was underlain by a very substantial large pointed post (see photo below) that was 17 cm in diameter and extended down as much as 75 cm deep. Its function however, remains unknown. A virtually identical feature and large underlying post was also uncovered in Block Unit B to the north. There it cut through the large Feature #5 house pit feature deposits and so, must date after ca. 3000 BP.


Profile of large pointed post of Terminal Archaic age located under Feature #36.


Our work at Davidson, because it is a research project, has allowed us the luxury of examining a large Archaic site over several years including the ability to do multiple surface collections and mapping at a scale never before attempted on Ontario sites of this age. We have also been able to apply techniques such as gradiometer surveying at a scale never before attempted on such Ontario sites. The result has been the thorough documentation of a large, 1.9 ha Late Archaic site that has significant Broad Point Archaic and Terminal Archaic Small Point associated occupations, dated as a whole to between 3800 and 2800 BP years ago. Broad Point material ocurs over the whole site area and such large sites seem to be characteristic of many Broad Point ones in the vicinity and much larger than almost all Archaic sites. Preliminary analyses suggest the Broad Point material occurs in several clusters of about the same size spaced roughly equally across the site, leading us to suspect the site suspect the site may have been used as an aggregation site at that time where, over a lengthy period, often dispersed peoples came together seasonally to share information and resources. However, these large sites are not typical of the Small Point Archaic ones known anywhere and indeed, Small Point sites in similar situations (e.g. in littoral areas adjacent to extensive wetlands bordering the Great Lakes) are usually smaller and more ephemeral (see Ellis et al. 1990:114, 2009:821-822).

Significantly, one extensive area of the Davidson site, covering at least 3500 m2, consists of buried, fully or partially (e.g. ploughing has only cut into the top of the buried humic layer) undisturbed deposits containing large amounts of artifactual debris and numerous cultural features of types never before seen on Ontario sites of this antiquity. Notable features include substantial true middens and storage pits. Sealed, undisturbed deposits, and especially of this age, are incredibly rare. Moreover, the deposits contain preserved organic remains that are rarely found on these ancient sites and that have allowed us to provide refined age estimates of this material via radiocarbon dating (and superposition of features), allow us to document some aspects of the rarely preserved organic (bone/antler) tool kit, and provide direct evidence of the subsistence and settlement practices of these early occupants of the province.

It is possible to argue that the seasonal use of the site varied over the time of the occupations. For example, the Broadpoint component definitely includes a large storage pit that suggests based on its contents of nutshell, turtle shell, fish bone, berry seeds and so on, suggest at least a summer to fall occupation -- this interpretation is consistent with the evidence collected by Ian Kenyon (1979a) for use of the site in Broad Point times. In-ground storage as indicated by the large pit Feature #1 also suggests winter abandonment in Broad Point times. Based on DeBoer’s (1988) ethnographic study, in-ground storage amongst mobile groups was done by those who wanted to protect the contents from animals or sometimes, hide from other humans, as they abandoned sites in winter. Such a strategy was practiced in more recent times by native peoples of the region (e.g. Holman and Krist 2001:9). In other words, and somewhat counter-intuitively, in-ground storage pits are actually indicative of people abandoning sites for significant periods. Resources were accumulated and stored in pits that could be accessed by visits in the rest of the year at other seasons. This practice has been referred to as “dispersed caching” and it is interpreted primarily as an insurance or backup storage strategy (Morgan 2012:716-718)  – people would only return periodically or in times of food shortage (such as early spring) to access the stored pit contents.  

In direct contrast, the pit houses at the site date after 3200 BP or to Small Point/Terminal Archaic times. Pithouses are definitively colder weather structures, used in the fall to winter, due to their greater insulating properties (see Gilman 1987:541-542). Of some significance, not all groups used them in the winter. Notably those groups that practiced winter mobility or moved their camps in the winter did not build such houses. Only groups who did not move throughout much of the cold seasons were willing to invest the considerable labour needed to build pit houses and Gilman (1987:542-544) demonstrates that those who were less mobile and built semi-subterranean structures often relied heavily on food storage to overcome local resource shortcomings and allow them to remain in one place for extended periods -- but as they remained on site they were storing above groudn so in groudn stroag features are lacking at that time. Regardless, Davidson stands out as unusual because of the large number of substantial features that are represented as compared to other investigated Archaic sites and I would suggest that is because sites investigated to date have not had cold weather, and particularly winter, occupation components where substantial housing and storage facilities are more likely to occur. Occupations in the fall to winter contradict directly the long-standing idea that Terminal Archaic peoples occupied littoral areas in the warmer months and moved into the interior well away from the lake shores/associated marshlands in the winter months (e.g. Ellis et al. 1990:114).

It is plausible therefore to suggest that the early Terminal Archaic component has a fall to winter occupation(s). However, one Davidson structure is not a pit house but a surface house outlined by a wall trench (Feature #32). As that structure had a preserved floor but no internal hearth, it may in fact represent a warmer weather occupation such as in the summer through fall. For what it is worth, the comparable wall trench house from the Rock Hearth site in Michigan had, among others things, some fish, birds, beaver, muskrat and deer, abundant turtle, and nutshell was very common suggesting a Summer to Fall occupation (Garland et al. 1990:199, 213). The wall trench house at Davidson is stratigraphically placed intermediate in time between the two styles of Terminal Archaic pit houses as the site, suggesting perhaps seasonal use changes over time during that general era. Overall, the suggestion of seasonal shifts in the use of the Davidson locality are not accompanied by any clear evidence of climatic environmental change suggesting they might not be understood in terms of adaptation to the local environmental conditions -- it may be there was a range of choices available to these Terminal Archaic peoples in terms of seasonal use of the locality and the choices taken were determined by more than simply local resource availability.

It is also possible, however, that our sample is biased by the small area excavated. It could be, for example, that the site was used for all or most of the year in the Terminal Archaic and that at the same location the inhabitants had two different types of houses for warm and cold weather use -- such alternative seasonal housing on the same sites does occur on later sites in some areas (e.g., Faulkner 1977; Rodning 2009:651).  One needs to at least consider that the site may have been used year round at the time and that these peoples were even more sedentary than anyone has suggested to date. In fact, a range of seeds from raspberry, plum and pin cherry from the late (post-3000 BP) Terminal Archaic Feature #22/64 fill does suggest warmer weather use at the same general time as winter pithouses were apparently in use at the site. At a minimum, the evidence suggests a level of settlement stability, including a more sedentary lifestyle (e.g. fewer residential moves) and consistent return seasonally to the same site locations year after year, than heretofore suspected (or to be more acurate, that has been easy to document; see Ellis et al. 2009) for this time period and notably in the Terminal Archaic. Certainly, these data suggest that the Terminal Archaic peoples may have been as "sedentary" as the later Middle Woodland peoples. The true midden and large storage pit associated with the Broad Point occupation also suggest decreasing mobility. Midden deposits in particular are clearly indicative of  organized refuse disposal as is characteristic of longer occupations.

Whatever the case, it is not an exaggeration to say that the Davidson site is amongst the most important pre-contact sites discovered to date in southern Ontario and certainly amongst the occupations of the preceramic era.

Acknowledgments:  I am especially grateful to Mr. Rick and Mrs. Marlene Davidson for allowing me to carry out work at the site. A very special thanks as well to Brian Deller for his advice and assistance throughout the project.

During the test excavations of 2006 Stan Wortner assisted in getting the site ready for excavation and Bob Pearce and the Museum of Ontario Archaeology loaned field equipment to carry out the work. Of course, I extend a special vote of thanks to the hard working 2006 volunteer field crew including: Darryl Dann, Chris Dalton, Elise Dalton, Jim "GIS" Keron, Bob Pearce, Steve Timmermans, Nancy Van Sas and Stan Wortner. Lisa Hodgetts assisted in identifying animal bones and Derek Paauw sorted and identified floral materials from the test excavations.

Darryl Dann, Ed Eastaugh, Toni Largo, Alanna Marsden, Cliff Patterson, Andrew Stewart and Krysta Terry helped me in mapping, surface collecting and doing soil probes at the site during the 2007 field season.

The 2008 to 2014 fieldwork at the site was funded by grants from the Social Science and Humanities Research Council of Canada. Additional funding used to purchase equipment (gradiometer, total station) was provided by an Academic Development Fund, University of Western Ontario, grant to Ellis, Lisa Hodgetts and Robert Pearce and the gradiometer provided by Sustainable Archaeology was purchased with a Canadian Foundation for Innovation project grant awarded to Neal Ferris and Aubrey Cannon. The regular field crew included Lindsay Foreman (Assistant Field Director), James Keron (Associate Field Director), Nancy Van Sas(Assistant Field Director), Nathalia Acosta (University of Western Ontario), Colin Bressette (Kettle and Stony Point First Nation), Darryl Dann (volunteer), Jennifer George (Kettle and Stony Point First Nation), Travis George (Kettle and Stony Point First Nation), Jenna Jaworski (University of Western Ontario), Lindi Masur (University of Western Ontario), Steve Naftel (University of Western Ontario), Larry Nielsen (volunteer and University of Western Ontario), Matt Seguin (McMaster University) and semi-regular Chris Dalton (volunteer). Many other volunteers also contributed to the success of the project including Rick Baskey, Matt Beaudoin, Eliza Brandy, Abigail Dalton, Justine Dalton, Lorraine George, Sue Haryatt, Sherman Horn, Thorsten Kragh, Ernest Lara, Alexandra Lausanne, Brenda Lausanne, Bevan Lindsay, Kaitlyn Malleau, Lindi Masur, Katie Mather, Dana Millson, Miggs Morris, Monica Norris, Jesse Oliver, Lisa Paulaharju, Harvey Reid, Jan Vicars, Gary Warrick, Deb Wilson and Stan Wortner. Other fieldworkers included summer works students made available through the auspices of Shari Prowse, Ontario Ministry of Culture, Heritage Branch. These included:  Kaye Lynn Boucher, Melissa Edwards, Kathryn Gilbert, Eryn Holborn, Allison Knott, Kelsey McLellan, Marilyn Varney, Kristen Waldick, Jennifer Watt and Dennis Wilson.

Logistical and other assistance was kindly provided by Greg George of Kettle and Stony Point First Nation, Pete Timmins, Kim Clark and Fernando Larrea and Brian Deller. Ed Eastaugh carried out the gradiometer survey and Jean-Francois Millare assisted with mapping. Ed and Jim Keron also generated many maps and drawings found in this report. Analyses of soil samples, and advice on interpreting site geology, were provided by Andrew Stewart, Joseph Desloges, John Menzies and Roger Phillips.  Stephen Moncton is analyzing the recovered plant remains and Lisa Hodgetts and Lindsay Foreman have provided advice on faunal identifications. Finally, several people have been of immense aid by washing and cataloguing the large number of artifact recoveries. These have included: Monica Blaylock, Rebecca Anne Dillon, Marissa Evers, Catherine Finan, Lindsay Foreman, Hannah Graves, Pam-Marie Guzzo, Lindi Masur, Laurie-Lynn Percy, Claire Venet-Rogers, Rammi Wickramasuriya, and those ever present stalwarts, Larry Nielsen and Darryl Dann.



References Cited

Ames, Kenneth and Hervert Maschner (1999):  People of the Northwest Coast: Their Archaeology and Prehistory. Thames and Hudson, London, U.K

Bettarel, Robert L. and Hale G. Smith (1973): The Mocassin Bluff Site and the Woodland Cultures of Southwestern Michigan. Anthropological Papers, Museum of Anthropology, University of Michigan No. 49.

Binford, Lewis R., Sally R. Binford, Robert Whallon and Margaret Ann Hardin (1970): Archaeology at Hatchery West. Memoirs of the Society for American Archaeology 24.

DeBoer, Warren (1988):  Subterranean Storage and the Organization of Surplus: The View from Eastern North America. Southeastern Archaeology 7:1-20.

Eastaugh, Edward, Christopher Ellis, Lisa M. Hodgetts and James R. Keron (2013): Problem-Based Magnetometer Survey at the Late Archaic Davidson Site (AhHk-54) in Southwestern Ontario. Canadian Journal of Archaeology 37(2):274-301

Ellis, Christopher (2006): A Preliminary Report on the 2006 Test Excavations at the Davidson Site: An Archaic "Broad Point" Component. Kewa 06(7):1-16.

Ellis, Christopher (2010): The 2008 Davidson Site (AhHk-54) Excavation Project. (125 pages) Ontario Archaeological Licence Report on File (P004-009-2008), Ontario Ministry of Culture, Heritage and Libraries Branch, Toronto, Ontario.

Ellis, Christopher (2011): The 2009 Excavations at the Davidson Late Archaic Site (AhHk-54) Ontario. (ii + 397 pages). Archaeological Licence Report on File (P004-011-2009), Ontario Ministry of Tourism and Culture, Heritage and Libraries Branch, Toronto, Ontario.

Ellis, Christopher (2013): The 2011 Surface Collections/Mapping and Gradiometer Survey at the Davidson Late Archaic Site (AhHk-54), Ontario. (i + 18 pages). Archaeological Licence Report on File (P004-015-2011). Ontario Ministry of Tourism and Culture, Heritage and Libraries Branch, Toronto, Ontario.

Ellis, Christopher (2014): The 2012 Surface Collections/Mapping and Gradiometer Surveys at the Davidson Late Archaic Site (AhHk-54), Ontario. (ii + 28 pages). Archaeological Licence Report on File (P004-016-2012). Ontario Ministry of Tourism and Culture, Heritage and Libraries Branch, Toronto, Ontario.

Ellis, Christopher and Edward Eastaugh (2014): Surface Collections/Mapping and Geophysical Surveys at the Davidson Late Archaic Site (AhHk-54), Ontario. (ii + 22 pages). Archaeological Licence Report on file (P004-018-2013). Ontario Ministry of Tourism, Culture and Sport, Heritage and Libraries Branch, Toronto, Ontario.

Ellis, Christopher, Edward Eastaugh and James Keron (2010): A Preliminary Report on a Late Archaic Pithouse from the Davidson Site (AhHk-54). Kewa 10(6-7):1-12.

Ellis, Christopher, Edward Eastaugh, James Keron and Lindsay Foreman (2006): A Preliminary Overview of the 2008 Excavations at the Davidson (AhHk-54) ‘Broad Point’ Archaic Site. Kewa 09(1-2):1-19.

Ellis, Christopher, Ian T. Kenyon and Michael W. Spence (1990): The Archaic. In: The Archaeology of Southern Ontario to AD 1650, edited by Christopher Ellis and Neal Ferris, pp. 65-124. Occasional Publications of the London Chapter, OAS No. 5. London, Ontario.

Ellis, Christopher and James Keron (2010): A Preliminary Report on a 3000 Year Old “Wall Trench” Structure from the Davidson Site (AhHk-54). Kewa 11(2):1-10.

Ellis, Christopher, James R. Keron, John H. Menzies, Stephen G. Monckton and Andrew Stewart (2015): For Immediate Occupancy: Cosy 3000 Year Old Heritage Winter House with River View Near Lake Huron. Apply to Terminal Archaic Realty. In Building the Past: Studies of Prehistoric Wooden Post Architecture in the Ohio Valley-Great Lakes Region, edited by Brian G. Redmond and Robert A. Genheimer, pp. 29-62. University Presses of Florida, Gainesville, in press.

Ellis, Christopher, Peter A. Timmins and Holly Martelle (2009): At the Crossroads and Periphery: The Archaic Archaeological Record of Southern Ontario. In: Archaic Societies of the Midcontinent, edited by Thomas E. Emerson, Andrew C. Fortier and Dale McElrath, pp. 787-839. State University of New York Press, Albany.

Emerson, Thomas and Dale McElrath (2009): The Eastern Archaic and the Tyranny of Theory. In: Archaic Societies of the Midcontinent, edited by Thomas E. Emerson, Andrew C. Fortier and Dale McElrath, pp. 23-38. State University of New York Press, Albany.

Faulkner, Charles (1977): The Winter House: An Early Southeast Tradition. Midcontinental Journal of Archaeology 2:141-159.

Ferris, Neal, and Michael W. Spence (1995): The Woodland Traditions in Southern Ontario. Revista del Arqueologia (Journal of American Archaeology) 9:83-138.

Fisher, Jacqueline A. (1987): Brodie and Parkhill: An Analysis of Two Late Archaic Broadpoint Samples from Southwestern Ontario. Kewa 87(8):3-21.

Fisher, Jacqueline A. (1997): The Adder Orchard Site: Lithic Technology and Spatial Organization in the Broadpoint Late Archaic. Occasional Publications of the London Chapter, OAS No. 3. London, Ontario.

Garland, Elizabeth, Caven P. Clark, Michael J. Higgins and Kathryn E. Parker (1990): The Rock Hearth Site (20BE306). In: Late Archaic and Early Woodland Adaptation in the Lower St. Joseph River Valley, Berrien County, Michigan,  pp. 173-236. Michigan Cultural Resource Investigation Series, Volume 2. Michigan Dept. of Transportation, Kalamazoo, Michigan.

Gilman, Patricia (1987): Architecture as Artifact: Pit Structures and Pueblos in the American Southwest. American Antiquity 52:538-564.

Hayden, Brian (1997): The Pithouses of Keatley Creek. Harcourt Brace College Publishers, Fort Worth, Texas.

Holdaway, Simon and LuAnn Wandsnider (2006): Temporal Scales and Archaeological Landscapes from the Eastern Desert of Australia and Intermontane North America. In Confronting Scale in Archaeology: Issues of Theory and Practice, edited by Gary Lock and Brian Molyneaux, pp. 183-201. Kluwer-Plenum Press, New York.

Holman, Margaret B. and Frank Krist Jr. (2001): Late Woodland Storage and Mobility in Western Lower Michigan. The Wisconsin Archaeologist 82(1-2):7-32.

Kenyon, Ian T. (1978): The 1977 Salvage Excavations at the George Davidson Site (AhHk-54). Ontario Ministry of Culture and Recreation, Historical Planning and Research Branch, Southwestern Region, Conservation Archaeology Report 1. London, Ontario.

Kenyon, Ian T. (1979): The 1978 Salvage Excavations at the George Davidson Site (AhHk-54). Ontario Ministry of Culture and Recreation, Historical Planning and Research Branch, Southwestern Region, Conservation Archaeology Report 3. London, Ontario.

Kenyon, Ian T. (1980a): The Satchell Complex in Ontario: A Perspective from the Ausable Valley. Ontario Archaeology 34: 17-43.

Kenyon, Ian T. (1980b): The George Davidson Site: An Archaic 'Broad Point' Component in Southwestern Ontario. Archaeology of Eastern North America 8: 11-28.

Kenyon, Ian T. and William Fox (1983): The Wyoming Rapids Saugeen Component: 1983 Investigation. Kewa 83(7):2-10. 

Lovis, William A., Kathryn C. Egan-Bruhy, Beverly A. Smith and G. William Monaghan (2001): Wetlands and Emergent Horticultural Economies in the Upper Great Lakes: A New Perspective from the Shultz Site. American Antiquity 66: 615-632.

MacDonald Robert I. and Ronald Williamson (1997): Area 4 – Truck Yard, Truck Pads, Infrastructure and Landscaping. In In the Shadow of the Bridge, The Archaeology of the Peace Bridge Site (AfGr-9), 1994-1996 Investigations, edited by R. F. Williamson and R. I. MacDonald, pp. 201-321. Occasional Publications of Archaeological Services Inc., Volume 1. Toronto, Ontario.

Monaghan, G. William and William A. Lovis (2005): Modeling Archaeological Site Burial in Southern Michigan. Michigan State University Press, East Lansing, Michigan.

Morgan, Christopher (2012):  Modeling Modes of Hunter-Gatherer Food Storage. American Antiquity 77:714-736.

Robertson, James A. (1987): Inter-Assemblage Variability and Hunter-Gatherer Settlement Systems: A Perspective from the Saginaw Valley, Michigan. PhD Dissertation, Department of Anthropology, Michigan State University, east Lansing, Michigan. University Microfilms Inc., Ann Arbor, Michigan.

Rodning, Christopher B. (2009): Mounds, Myths and Cherokee Townhouses in Southwestern North Carolina. American Antiquity 74:627-663.

Sassaman, Kenneth (2010): The Eastern Archaic Historicized. Altamira Press, Lanham, Maryland.

Sassaman, Kenneth (2011): History and Alterity in the Eastern Archaic. In: Hunter-Gatherer Archaeology as Historical Process, edited by Kenneth Sassaman and Donald H. Holly Jr., pp. 187-208. The University of Arizona Press, Tucson.

Springer, Chris and Dana Lepofsky (2011): Pithouses and People: Social Identity and Pithouses in the Harrison River Valley of Southwestern British Columbia. Canadian Journal of Archaeology 35:18-54.

Wandsnider, LuAnn (2008): Time-Averaged Deposits and Multitemporal Processes in the Wyoming Basin, Intermontane North America. In Time in Archaeology: Time Perspectivism Revisted, edited by Simon Holloway and LuAnn Wandsnider, pp. 61-93. University of Utah Press, Salt Lake.

Wilson, James A. (1990): The Boresma Site: A Middle Woodland Base Camp in the Thames River Valley. MA Thesis, Dept. Of Anthropology, McMaster University, Hamilton, Ontario.

Wilson, James A. (1991): A Bad Analogy? Northern Algonquian Models and the Middle Woodland Occupations of Southern Ontario. Kewa 91(4)9-22.

Woodley, Philip (1990): The Thistle Hill Site and Late Archaic Adaptations. Occasional Papers in Northeastern Archaeology No. 4. Copetown Press, Dundas, Ontario.

Back to Chris' Archaic page

Back to Chris' home page

Page last updated: May 12, 2015. CJE.