«By James R. Keron Graduate Program in Anthropology Submitted in partial fulfillment of the requirements for the degree of Master of Arts Faculty of ...»
General Comments In considering the distribution of chert types in general, examining all of the observations for chert source type, a number of generalizations can be made. The first point is that the western portions of the site have lower densities of detritus and therefore do not show up as statistically significant for a number of categories. This includes areas 4, 5, and 7 in particular. Second, it is clear that Kettle Point chert occurs with a higher frequency in Areas 2 and 3 and their middens. Third, on the other hand, Onondaga chert seems to be evenly distributed except that it tends to not occur in Midden 1 outside the palisade. And fourth, in some of the other areas of the site the lack of Kettle Point chert seems to be compensated for by increased use of local till chert.
In order to look closer at these generalizations, a further analysis was done where the site was divided into two regions, one a combination of Middens 2 and 3 and their associated areas and the other with everything else. This analysis clearly indicated that the frequency of Kettle Point chert in these areas is higher than the rest of the site and that the results are statistically significant. Also, local till chert has a higher frequency in the rest of the site and the differences are statistically significant. Using the same split of the site, the difference in the frequency of Onondaga chert was not significant.
11. Dividing the site into two areas the various chert percentages are as follows $ In the Midden 2-3 area the percentage of Kettle Point chert is 22.2 +/while the rest of the site has 11.9 +/- 4%. The differences are
Next, considering the distribution of Onondaga chert, another split of the site was created that also divided the site into two zones. One zone included Midden 1 and the adjoining portion of Area 1 that lies outside the palisade, and the other was the rest of the site assumed to be inside the palisade. This confirmed the earlier observation that Onondaga chert occurs more frequently inside the village than in Midden 1 and the difference is statistically significant.
While that completes the chert source analysis that is statistically significant, mention should be made of Area 7 since the various frequencies are distinctly different from the rest of the site. Unfortunately, the sample is low so there are no statistically significant differences excepting that local till chert is very low compared to the rest of the site. The flakes in this area are the higher quality Kettle Point, Onondaga and other chert. Also, looking at the flake types, most of these are either bifacial retouch (25%) or fragmentary flakes (50%), both of which are suggestive of late stage reduction. While the differences are not significant, it is suggestive that the knapping activity taking place in Area 7 was different than the rest of the site.
Another area of the site that is unique is Midden 1. During excavation for the UWO field school (Ellis 1996), it was determined that this midden was outside the palisade. As noted above, the midden has a lower frequency of Onondaga chert. Looking at the flake types we also note that there is a lower frequency of decortication flakes than that found elsewhere in the village. This is especially puzzling as the midden also has a high frequency of local till chert and in general local till chert has a higher percentage of decortication flakes than do the other chert source types over the site. This suggests that the reduction and use cycle leading to deposition of debitage in Midden 1 was different than that for the other middens in the site. Furthermore, while the CSP shows a frequency for Kettle Point chert in the 10% range, personal observation of flake samples from the midden excavation indicate that the actual frequency is much lower than that.
However, to support this observation would require a detailed analysis of the debitage.
In considering the flake types, the purpose of the spatial analysis is to understand the general pattern of the reduction sequence in the village. As noted above in the observations, the differences in the frequencies of the various flake types between the various areas is weak. Two types, bifacial retouch flakes and shatter, show no significant differences between the site areas and another, decortication flakes has been discussed above for Midden 1. In looking at flake fragments and core trimming flakes, Area 3 shows differences in both of them. Going back to the chert source types, Area 3 appears to be higher than the rest of the site in both Kettle Point chert (24.1%) and Onondaga chert (27.8%). Taken together this would suggest that the reduction activities taking place in Area 3 and the associated midden were different than that for the rest of the site.
The residents had better access to higher quality chert and the reduction processes were different.
Cassandra Site Observations The Cassandra site is a small late prehistoric Neutral village or hamlet located just north of Lambeth along the same creek on which the Thomas Powerline site is found. A CSP was conducted by the author in 1998 (Keron 2000) that clearly identified the site boundaries and three middens within the site. As there are clearly defined middens, the cultural areas defined for the site were based on these three middens and as before, the rest of the site was divided into three other areas based on proximity to the nearest midden. The cultural areas are identified in Figure 8. The data used to develop this analysis are presented in Appendix C: Tables C6 through C8.
Chert Source Type Analysis The initial analysis was conducted to examine the distribution of unidentified chert yielding the following observation.
1. There is no statistically significant spatial patterning of unidentified chert.
The next spatial analysis was conducted on Kettle Point chert, Onondaga chert and local till chert as measured against the six cultural areas shown in Figure 8. The analysis was not attempted on Aother@ chert as the numbers were too small. The following are the observations.
2. The difference in percentages of local till chert between Midden 2 and Midden 3 is different with statistical significance.
3. Kettle Point chert is the most commonly used chert, forming over half the sample in each area. The frequencies across the site are similar in the 50-60% range.
4. Midden 3 has the highest percentage of Onondaga chert but the differences from other areas are not quite statistically significant.
5. Middens 1 and 2 do not have the same access to Onondaga and compensate with increased use of local till chert.
Given the distribution of patterning of Onondaga and local till chert, a second round of analysis was conducted where the number of areas was collapsed from six to three by combining the associated middens and areas. The analysis in this manner yielded the following observations.
6. The differences between Area 2 and Area 3 are significant for local till chert.
7. The differences between the same areas for Onondaga are significant at about the 90% (rather than the normally used 95%) level.
Flake Type Analysis.
The analysis of the distribution of the various flake types was conducted with both the six areas and the three areas with the following result.
8. None of the flake types have any significant distribution. Midden 2 has a higher frequency of core trimming flakes but the difference is not significant.
General Comments In general at the Cassandra site, the most commonly used chert is Kettle Point chert and all areas of the site have equal access to it. One area around Midden 3 has preferred access to Onondaga and the rest of the site does not have the same access to this chert. The rest of the site compensates for this lack by use of local till chert. All reduction activities take place throughout the entire site.
Dorchester Site Observations The Dorchester Village (AfHg-24) is a MOI village belonging to the Uren Substage. It is located within two kilometres of the Calvert Site (AfHg-1) and the Mustos Site (AfHg-2). In 1998, the author with the help of Shari Prowse conducted a CSP on the site that indicated that the site is 1.5 hectares in size. It is the latest known village within the Dorchester cluster of sites. Unfortunately, no obvious middens or cultural areas were defined by the CSP. Perhaps refuse disposal was still handled in the Glen Meyer fashion or perhaps it was thrown over the edge of the steep slope to the north creating as yet unlocated hillside middens. In any event the interior spatial analysis was forced to proceed by trial and error, primarily by dividing the site into thirds and halves as described in the following analysis. Figure 9 shows the various areas discussed in the analysis. The data used to develop this analysis is presented in Appendix C: Tables C9 through C12.
Chert Source Type Analysis In lieu of any visible middens around which to define cultural areas, the first round of the analysis was done using the three arbitrarily defined areas shown in Figure 9.
While this is certainly much less than desirable, an approach like this would have eventually discovered the major trends in Brian and Cassandra. The observations are as follows.
1. Local till chert is evenly distributed over the site.
2. Onondaga and Kettle Point cherts are evenly distributed in the central and eastern areas.
3. In the western area, Kettle Point chert occurs with a higher frequency and the frequency of Onondaga chert is less frequent.
4. However none of these trends is statistically significant.
A second analysis was run where the central and eastern areas were combined and compared to the western area. However, while the same trends were apparent, none were statistically significant. The map of the distribution of Kettle Point chert was then examined and there appeared to be some patterning to it. First, there was a background distribution of Kettle Point chert over the site that was fairly even. However, there was one spot where half of the Kettle Point chert flakes grouped together in a fairly tight cluster. Accordingly, this area was circled (Figure 9) and another two-area map produced that contained the cluster in one area and everything else in the other. This definition of cultural areas was then run through the analysis with the following results.
5. Local Till chert is identical between the two areas.
6. In the small area to the west, the frequency of Kettle Point chert is much higher than the rest of the site, 31.6% versus 5%.
7. Onondaga is significantly lower in that same area with statistical significance.
Flake Type Analysis The flake types analysis was run through several different cultural areas.
8. There were no statistically significant differences in the distribution of the various flake types.
General Comments The patterning within the Dorchester Village site is similar to that observed for the earlier sites with the exception that the various chert types have reversed their position. The chert type that is evenly distributed is local till chert. There is differential but restricted access to Kettle Point chert within one small zone of the site. In the other zones where access to Kettle Point chert is severely restricted, lack of access is compensated for by a greater use of Onondaga chert.
Mustos Site Observations The Mustos Site (AfHg-2) is a Glen Meyer Branch village located in the Dorchester cluster. It is approximately 2 km southwest of the Dorchester Village site (AfHg-24). The Calvert Village (Timmins 1997), also a Glen Meyer Branch village, is about 400 m to the northeast of Mustos. A CSP was conducted by the author (Keron
1986) with the aid of Peter Timmins and this material forms the basis of the following analysis. Again, there are no clearly defined middens that could be used as the basis for defining cultural areas within the site so, as with the previous analysis, it was necessary to again proceed by trial and error. The cultural areas used in the analysis are illustrated in Figure 10. The data used to develop this analysis is presented in Appendix C: Tables C13 through C17.
Chert Source Type Analysis
1. Using the east-west division of the site there are no differences that are significant.
2. Using a north-south division, Kettle Point chert frequency is higher in the middle band and lower in the south band. Local till chert is the reverse but the differences are not significant.
3. There is no Kettle Point chert in the little cluster to the north.
Flake Type Analysis
4. Using the east-west division, shatter is more apt to be found in the eastern side of the site (17.2% versus 6.3%).
5. Bifacial retouch flakes are more apt to be found on the western side of the site but the differences are not significant.
6. Using the north-south division, bifacial retouch flakes are more apt to be found in the middle section of the site than in the southern section (10.7% vs 2.7%) but the difference is not statistically significant.
General Comments Despite trying several divisions of the site there do not appear to be any significant differences in the distribution of chert source types over the site. Looking at the debitage types there is a tendency for shatter to occur towards the southeast corner of the site. This last trend is especially puzzling since, in theory, shatter and decortication flakes should tend to co-vary especially where local till chert is involved. Unfortunately, with the Mustos sample, much of the shatter is unidentifiable as to source. The local till chert debitage that was either shatter or decortication was plotted but this did not follow the same clustering to the southeast pattern.
In both the east-west and north-south analysis there was some patterning of bifacial retouch flakes. In fact, looking at their distribution there appears to be two loci where they are located, one in the west central portion of the site and another represented by two flakes in the southeast corner.