«By James R. Keron Graduate Program in Anthropology Submitted in partial fulfillment of the requirements for the degree of Master of Arts Faculty of ...»
Glen Meyer Pattern The linear nature of the Glen Meyer distance decay was evident looking at the plot of the sites (Figure 12) before even the regression line was added. There are however, three significant outliers that demand examination. The McGrath site (AfHh-61) in the Byron area (Poulton 1985a) is notable for an extremely high percentage of Kettle Point chert in the debitage at 99.1%. The formal artefacts were largely Kettle Point chert at 57.9% but a significant number were either Onondaga and/or local till chert at a combined total of 28.1%. This site was excavated as part of a salvage project and no typical Glen Meyer settlement data (e.g. houses palisades etc.) were evident. Poulton (1985a) interprets the site as a small temporary camp. This observation is important as it implies a very restricted duration of use that is different than Glen Meyer villages sites such as Calvert where Timmins (1997) estimates successive occupations covering 60 or more years. Larger villages are subject to an averaging effect over time whereas the McGrath site is a short time capsule representing one episode in chert acquisition and use.
In all probability, this site represents a small temporary camp of a group who had recently visited Kettle Point and were returning with a supply of chert. This pattern, where the debitage has a higher percentage of one chert than the discarded artefacts, has been noted in earlier times (Ellis and Spence 1997) and is there interpreted as indicating that the debitage represents the most recent source to be visited by more mobile groups.
The other two Adeviant@ EOI sites are exactly the opposite as the percentage of Kettle Point chert is very low. Both of these, Melbourne-7 (AfHj-17) and Caradoc-3 (AfHj-105), are part of the Caradoc cluster of sites and both are known only from small surface collections. However, both are at the small end of the site sizes reported by Williamson (1985) so a briefer occupation is quite possible. Also, it should be noted that Melbourne-7 is the most southerly of the reported Glen Meyer sites in the Caradoc cluster and so could quite likely be a stop on the way back from the Lake Erie shoreline after chert acquisition. While not evident in the site data in Appendix F, one of the qualitative notes from this site indicated that the Onondaga chert debitage was larger and of better quality than that typical of most Iroquoian sites. The larger size could well be the indicative of the recent acquisition.
With respect to the specific chert working industry, Morrow and Jeffries (1989) postulated that there should be differences in chert use if the more distant chert was acquired through embedded procurement as opposed to exchange or special purpose acquisition trips. The observation 7 in the section on artifact variation in chapter 6 notes the fact that all chert types seem to occur in the same percentage as the discarded artefacts suggesting that the more distant cherts were not treated any differently than local chert, thus, implying embedded procurement of both Onondaga and Kettle Point chert.
In examining the regression line, with the exception of the three outliers discussed above, it is clearly best approximated by a linear equation and the R-squared value is high indicating a good fit. It is also very different from the curve that arises later, probably during the MOI. In reviewing the distance decay literature, regression lines where the best fit is linear are always associated with direct procurement (Findlow and Bolognese 1982; Hodder 1974; Renfrew 1977; Torrence 1986). The normal line of reasoning is that the cost of acquisition is driven solely by the cost of the trip to obtain the material. The other point that Renfrew makes is that within the supply zone of a down-the-line exchange system, everyone has access to the material and that the frequency declines only slightly with distance to edge of the supply zone. This edge is usually interpreted as the boundary with a neighbouring corporate group. If this really is the case, then extending this analysis into EOI groups in the Catfish Creek and Norfolk communities could prove interesting.
One final result that is also important to this interpretation is that a similar result has been obtained from the American Southwest. There, Findlow and Bolognese (1982) analysed the movement of obsidian from various sources in New Mexico while controlling for time using various Basketmaker and Pueblo periods. One source, Antelope Wells, shows a similar change from an earlier linear equation to an exponential equation in later periods that is associated with increasing social complexity. They interpret the earlier mode to be direct acquisition, which is replaced by a down-the-line exchange system as the socio-cultural complexity increases, a situation that is matched by
the results found here.
Hypothesis 3: Kettle Point Moved West to East The distance decay analysis of Kettle Point chert indicates that this hypothesis was the case for all time periods although different decay functions were evident. All that changes is the amount and the rate at which it decays with distance.
Hypothesis 4: Onondaga Moved East to West As was demonstrated in the last chapter, deriving a distance decay pattern for Onondaga chert proved problematic. The original assumption, that the chert was derived from the primary sources located east of Port Dover and passed from group to group overland, was clearly wrong. If this situation was the case, then a distance decay function would have been evident. Reid (1986) notes that a flat distance decay is a possibility but the only potential example was during Early Paleoindian times where Ellis (1989) notes a preference for use of high quality chert from a single source and suggests that the source used served as a band or social marker. It would seem highly unlikely that a flat distance decay function would exist in the Late Woodland period.
Given the difficulty in separating some forms of local till chert from Onondaga as discussed above and elsewhere (Keron 2003), there is still the possibility that the material classified as Onondaga is nothing more than local high quality till chert. There are, however, several arguments against this possibility. First, during fieldwork in the study area, the author has had the opportunity to observe a great deal of till chert both on and off Iroquoian sites. While there are occasionally small pebbles that are close to appearing Onondaga-like, no large nodules have ever been observed that are not in an archaeological context. Second, as was shown in the analysis in the previous chapter, the material classified as Onondaga is arriving in a more reduced form than even Kettle Point chert which in turn arrives on the site in a more reduced form than local till chert. Of the three primary types, Onondaga has less cortex and shatter than the other two chert types.
If the material identified as Onondaga chert was really just a variety of local till chert then it should be treated in the same way and one should see only expedient use (Andrefsky 1994). While the entire Iroquoian industry would be characterized as expedient, Onondaga chert is clearly treated differently than local till chert.
The other possibility, and the most probable, is a Lake Erie source for the Onondaga chert observed in the study area. Two explanations can be suggested. First, chert from the primary outcrops east of Port Dover could have been transported along Lake Erie by canoe. Fox (1990) has clearly demonstrated that this transport was occurring in the Lake Huron basin by the Odawa and links the traded material to several outcrops that are at or close to the shoreline. Potentially arguing against this explanation would be the historical observation that the Neutral lacked the means to transport beaver pelts to the French (Trigger 1987). However, the context of that statement is caught up in Native politics where the Huron were striving to maintain their role as middle-men and were actively seeking to discourage direct trade with the French by any of their upstream partners. It seems unreasonable to expect that while the Huron and New York State Iroquois were certainly adept at water transport, the Neutral were not. The second potential source of Onondaga chert from the south would be secondary deposits that can be found there. These have been discussed above with respect to the chert sources.
Looking at the Neutral period where down-the-line exchange has been established there appears to be one community between Neutral sites and the Lake Erie shore so it can be expected that the Neutral sites in the study area would not show any distance decay between each other. They are all one step removed from the source. With the more free ranging Glen Meyer people, again all sites are very close to the same distance from the source so again there would be no distance decay between the study area sites taken as a unit. Finally, unlike the case for Kettle Point chert, there is no single point source and the Onondaga chert could be obtained at a number of locations along the lakeshore.
At this point in time the best explanation would be that Onondaga chert was acquired from the Lake Erie shoreline either directly from secondary deposits or by transport from the primary sources by canoe along the lakeshore. Therefore, no distance decay is evident owing to the orientation of the study area. This possibility should be considered the best explanation for the phenomena at the moment. It is certainly reasonable but testing with a revised research design, the addition of sites from Elgin County and some investigation into the nature of the reported Onondaga chert along the shoreline would be required for validation.
Hypothesis 5: Exchange Routes and Lineage Control The antiquity of trade has been discussed above so the question that remains is to determine how far back in time this can be demonstrated. Chapter 5 contained the exploratory analysis of distributions of both chert source types and flake types within five Iroquoian sites in the area, one Glen Meyer, one Uren, one Middleport and two Neutral. The results of this analysis mapped against the implications of this hypothesis are shown in Appendix E, Box 5. In general, there was good support for Neutral and Middleport sites, problematic support for the Uren site and no support for the Glen Meyer site.
For both the Brian and Drumholm sites, there were clear indications of internal patterning within the site with one area of the site that had differential access to Kettle Point chert. This conclusion is consistent with the results obtained earlier at the Harrietsville site (Keron 1986). The Cassandra site (AfHh-65) also has internal patterning but it is not as pronounced. However, this site is one of the those like Thomas Powerline (AfHh-3) that has a high percentage of Kettle Point chert that may obscure the overall internal distribution. This site is also best interpreted as a satellite hamlet in the sense given by Pearce (1996), or possibly a small village, so it may have been composed of a single lineage. However, one portion of the site is clearly given over to use of high quality chert.
The case of the Uren sub-stage Dorchester site is more problematic. First, the analysis was complicated by the fact that there were no obvious middens as is normal with Uren period sites (Warrick 2000). Several attempts at analysis by sectioning the site failed to disclose any significant patterning although one end of the site seemed to have a higher percentage of Kettle Point chert. Visual inspection indicated a concentration within one discrete area of the site and when that was singled out it was found to be significant statistically. Given the very low percentage of Kettle Point chert on the site, this could be incipient lineage-controlled exchange or it could also be the result of one individual who distinguished himself by acquiring an unusual and potentially high status chert type given the hypothesized hostile relations with the Western Basin people. Chert acquired under such conditions would reflect on the audacity of the warriors that obtained it.
During the Glen Meyer period, no significant evidence of internal patterning could be found. This result is not unexpected for several reasons. First, it is assumed that the EOI groups occupying individual sites formed the basis of the lineages in later more complex MOI and LOI villages (Timmins 1997). With a single lineage at a site, there should be no evidence of differential access so the absence is not surprising. Indeed, any variance in distribution would have argued against the hypothesis so potentially the failure of the Glen Meyer site to show this may be in the wrong column of Appendix E, Box 5. Second, further complicating the potential to determine spatial patterning from surface collections of Glen Meyer sites is the tendency to rebuild on the same site. This pattern is quite evident at the Calvert site (Timmins 1997) where four distinct overlapping occupations are present. Potentially the only internal spatial analysis that could be done would be similar to Timmins analysis where debitage from each of the occupations was separated out revealing varying access to Kettle Point chert through time. However, the variation between occupations was not widely different. And finally, with a direct access pattern established that shows only a shallow slope in the distance decay line, lineage control of an exchange route is highly unlikely.
Further evidence supportive of the lineage control of trade routes is the variation that occurs in the amount of Kettle Point chert used at cabin sites. The best example is around the Lawson site where three cabin sites vary from 8% to 24% to 65% Kettle Point chert (Pearce 1996). In the Pond Mills cluster two adjacent cabin sites, Skinner (Keron
1989) and Laidlaw (Peace 1996) have different frequencies. The sites could well both be hamlets of the Brian site and the frequencies are not unlike the have and have-not area within the Brian site.
A third confirmatory indication is the consistency with the demonstration of a down-the-line exchange mechanism. Assuming lineage control of trade routes, then a down-the-line exchange system should be evident and thus the results tend to confirm lineage ownership.