«By James R. Keron Graduate Program in Anthropology Submitted in partial fulfillment of the requirements for the degree of Master of Arts Faculty of ...»
7. Trade patterns between groups are stable through time. Groups in the same clusters should show the same relative percentages of chert source in all time periods. If this is true, then during the shift in the Middle Ontario Iroquoian stage from sandy soils to clay soils occurs, it should be possible to link the groups based on chert source percentage. The pattern previously observed in Pond Mills and Dorchester (Keron 1986) should be repeated elsewhere.
8. If the pattern of exchange has a considerable time depth and all other factors are equal, then it would be reasonable to expect that the pattern of use would be consistent through time. There should not be dramatic changes in the frequencies of the various flake types. A sharp break in a pattern of use would suggest that the time depth hypothesis does not hold true.
Exchange Medium Implications As discussed above, the hypothesis that the exchange medium was raw cores may be contradictory to some of the other hypotheses so in some sense this hypothesis is more exploratory. The following would be the measure of the possibility.
9. If cores are the primary medium of exchange, then there should be exhausted cores of all chert types in all sites. These exhausted cores should be at the same ratio as the debitage for the various chert types in each site.
10. Furthermore, the relative percentage of flake types within each chert type should stay the same regardless of the distance from the source. (i.e. The percentage of shatter and decortication should be similar in all sites).
Site Type Differences Especially in the late Iroquoian stage there are various types of sites varying from large villages to small agricultural cabin sites. It is presumed that different activities were carried out at these sites and that they were occupied for varying durations/seasons.
Therefore this should be reflected in the lithic detritus.
11. Embedded acquisition of local till chert at the cabin sites is a good probability since agricultural activities would result in a close scrutiny of much of the soil. There should be higher percentages of local till chert and more indications of core reduction such as higher amounts of decortication flakes and shatter.
12. As different activities occurred at these sites it could be hypothesized that the chert industry would also be different. Some trends were noted in the previous study but the results were equivocal so it is not possible to generate specific implications, so this would be best considered strictly an exploratory strategy.
Embedded Procurement of Local Till Chert Implications
13. Locally obtainable chert should be used in more expedient manners (Andrefsky 1994;
Morrow and Jeffries 1989). Local till chert will be used for more expedient flake tools and Onondaga and Kettle Point preferred for formal artefacts. For local till chert there should be more core trimming flakes and fewer bifacial retouch flakes and flake fragments.
This completes the list of implications and measures that could help establish the behavioural hypotheses outlined above. The data collected to evaluate these implications will be documented in the next chapter.
Confounding Hypotheses It would be wise to pay some attention to alternate hypotheses or factors that might invalidate the study, provide alternate explanations of the observed data or could indeed be considered alternative behavioural hypotheses. In some cases, it may be necessary to ensure that supportive data is collected during the study. The succeeding paragraphs consider some of these and the approach taken with each.
The dramatic difference between Lambeth and Pond Mills where the percentage of Kettle Point chert drops from over 60% to around 20% could be the result of some other factor besides inter-group exchange such as seasonality of occupation. At this point the best argument that can be made would be that there are permanent villages in each cluster as well as agricultural cabin sites so it is not unreasonable to expect that the differences in chert usage are a result of the acquisition options and distance. If one of the clusters was demonstrated to be lacking permanent villages then this argument might apply.
Hostile groups or other factors could disrupt the orderly flow of raw materials. One of the anomalies noted in the earlier study was the Uren Substage village in Dorchester that stood out in that it had a very low percentage of Kettle Point chert. At the time, potential explanations such as warfare with Western Basin people or potentially high lake levels were suggested as potential explanations. As the evidence was extremely weak, there was no point in attempting to determine causes. The intention of the comparison here is to determine if the trend noted at Dorchester is more widespread. The data to be collected will be capable of answering this question. If such a trend is established then determining the cause will be difficult, as the cause could well lie outside of the study area. As it relates to warfare, the number of projectile points on Iroquoian sites has been used as an inference of increased warfare (Finlayson 1998).
Water travel could mean that all costs are nearly identical since much larger quantities of chert could be returned than by land travel. Canoes make remote acquisition trips more profitable. This method would, in effect, lower the cost of chert acquisition resulting in more expedient use of the chert so acquired. If down-the-line exchange is suggested instead, this method of acquisition would mean that the communities downstream from the community acquiring chert through water transport would be excluded from water transport as a means of acquisition so this is not a significant concern. If this was a factor, then the anticipated impact on distance decay curves would be to flatten them out as Janusas (1984) has demonstrated with sites to the northeast of Kettle Point.
One of the implications of continuity through time is that it may be possible to trace individual groups through time based on their relative access to Kettle Point chert. Of course, this hypothesis assumes that the groups in the area remained stable through time.
Group migrations occurred particularly in the MOI stage and these could confound the model as the groups may relocate into areas previously occupied by other groups. Such movements could result in changes in chert usage percentage or the migrants could also develop alternate access to a common supplier so that the intrusive group would appear no different than the indigenous group. There is however, no way to plan for this possibility but it will need to be considered in the final discussion.
Another factor to be considered is that spurious indications of inter-group exchange might be created based on the physical properties of the different varieties of chert.
Kettle Point and Onondaga chert are both of higher quality and so could be used in either a curated or expedient fashion. Local till chert on the other hand is frequently not amenable to biface manufacture as the glacial action that deposited it quite frequently damaged the crystalline structure so that fracture planes and shatter are a frequent result of attempted reduction. Also some of it tends to be much coarser than the high grade sources and have non-siliceous content or flaws. Thus, a situation where it appeared that Kettle Point or Onondaga chert were preferred for bifaces could result from the physical nature of the stone not a cultural preference based on cost of acquisition as the theory might suggest. Although this generally low grade chert was largely ignored by earlier inhabitants (e.g. Late Archaic, see Ellis and Spence 1997) specifically because of the poor quality, it became a viable source with sedentary groups who seemed to adopt a greater use of expedient flake technology (Parry and Kelly 1987). The impact of this factor is that the demonstration of trade would be best identified through varying uses of the high grade sources rather than a comparison with the low grade sources.
There is some indication from the Lawson site (Pearce 1994) that, while the debitage is very high in Kettle Point chert, the projectile points are much more likely to be made from Onondaga chert. This pattern could be a result of the cost of obtaining Onondaga chert as described below, but it also could be derived from the basic properties of the chert itself. Flint knappers have described it as a Atough@ chert (Dan Long, personal communication). The data being collected should be able to distinguish between these two possibilities. If cost is the determining factor, then the corollary to high Onondaga in the west is high Kettle Point in the east. If this pattern holds up, then the selection is based on using high cost material as per the implications. If it does not hold up, then it may be more likely that the characteristics of the Onondaga chert made it more preferable for projectile points.
The alternative to down-the-line exchange, is that individual resource acquisition trips by specific individuals or embedded procurement by task groups is responsible for these patterns and no trade can be inferred. The data should be able to support or disprove this alternate hypothesis. If an embedded procurement strategy is used for all chert acquisition, there will be no difference in use between non-local chert and locally obtainable chert, (Morrow and Jeffries 1989) and all flake categories should be similar for each source. Another possible indication of this would be a regression curve with a linear equation as Reid (1986) and others associate this pattern with direct acquisition.
It is universally accepted that the shift from the EOI to the MOI entailed significant cultural changes. Indeed, Kapches (1995) declares that the advent of the MOI stage represents the establishment of the Iroquoian cultural pattern as recorded at the time of contact. While generally discounted by most archaeologists, the Pickering conquest hypothesis (Wright 1966) is still adamantly supported in some quarters (e.g. Finlayson 1998; Wright 1990,1992). Evaluation of the conquest hypothesis is certainly beyond the scope of this study. However, the nature of the study area provides a basis to assess the continuity of the organization of lithic technology over the transition from the EOI to the MOI stage.
Finally, there is always the possibility that some or most of the behaviourial hypotheses will not hold up on wider scrutiny. After all, the pattern was developed based on fewer sites from a smaller study area. So it is entirely possible that the perceived pattern was fortuitous and there is no real substance to it. If that happens then at least some spurious conclusions will have been removed from the literature but there may be opportunity for building an alternate acquisition hypothesis based on the data collected. In any event, that is a risk we all accept by conducting problem-oriented research and the result will at least further our understanding even if it results in disproving certain hypotheses. Obviously, no real data gathering plans can be based on this assumption.
Chapter 4: Data Collection Methodology
This chapter describes the methodology used to collect the data and provides definitions of the various information recorded. The first part of the chapter describes the chert sources, their geological setting, the problems involved in assigning a particular item to its appropriate source and a description of the controls used in attempting to maintain consistency. That is followed by the definitions of the various data collected and then by a discussion of issues surrounding analysis of debitage and various debitage typologies since the debitage typology is critical to this study.
Chert Sources: The Geological Setting In general, Ontario chert sources are well-known and well-documented both geologically and archaeologically (see Eley and Von Bitter 1989; Fox 1979e; Janusas 1984; Parker 1986b; Parkins 1977 ). In the study area specifically, there are no surface outcrops due to the thick mantle of secondary glacial deposits. As a result, there are no immediately available primary sources although there is a significant amount of secondary deposits in the glacial till. The closest primary source is Kettle Point chert located at Kettle Point, Ontario, 45 km from the west end of the study area. The other primary source used in measurable quantities is Onondaga chert which outcrops, starting at a point 56 km from the east end of the study area, and swinging southeast to the north shore of Lake Erie east of Long Point. Two other chert sources, Selkirk and Haldimand chert, are located just north of Long Point. These were used sparingly by the Iroquoian peoples in the study area.
The bedrock of southern Ontario is comprised of a series of beds of Devonian and Silurian age that, in general, slope gently downward to the west being the eastern edge of the Michigan Basin (Stokes et al. 1978). Along a line between Kettle Point and Niagara, the older beds occur to the east and these are subsequently covered by younger beds as one moves progressively westward (see Figure 6; Chapman and Putnam 1984). In the west, Kettle Point chert occurs at the interface between the Upper Devonian Ipperwash Formation limestone and the overlying Kettle Point Formation shale (Eley and Von Bitter 1989). Most of this primary chert source is just below water level offshore (Janusas
1984) but there is one outcrop a little to the northeast of the point proper that is right at the current water level. Chert can be obtained from the primary deposits but personal observations indicate that it is also found in the shallow water all around the point in sizes varying from small cobbles to large boulders containing sections of the chert bed.
Secondary deposits of Kettle Point chert occur southward into Lambton county at least as far as Shetland. However, the chert pebbles there are fairly small and of little use for knapping. Other secondary deposits can be found in the till at least as far east as the Thedford area (Janusas 1984).