Discussion
The above analysis represents a first pass on a single dimension of variability in material representations of the structure of meso- to macro-scale interaction among Scioto Hopewell households, communities, and aggregates thereof. While keeping in mind the limited scope of this analysis, it is apparent that for scales involving local and regional production, it is households and the relations among them that define Scioto Hopewell. Major earthworks, long thought to create and structure Hopewell interaction and integration and implicitly assumed by our dominant models to structure the interactions that form Hopewell, appear as emergent phenomenon from existing interaction between household-level actors.
Even at the Middle Scale where interaction is taking place across the multiple Hopewell societies of the Midcontinent, we see a significant representation of households. Yet at this scale we can see the beginnings of some differences - two slightly different forms of interaction take place; one preferentially featuring interaction to the west represented by resources from the Mississippi valley, and one preferentially featuring interaction to the southwest represented by resources from the lower Ohio valley. Time does not appear to be a factor in this distinction, as sites from both subgroups appear to cover the full range of Hopewell chronology (see Nolan et al. 2017, 2019). There also is a lack of geographic patterning that might indicate geographically delineated polities structuring the relations objectified by the distribution of these materials.
Production at this Middle Scale still seems to include household level actors in the interactions with neighboring societies. Yet earthwork sites seem to be increasingly important in the spatial structure of interaction across regional scales - particularly with regard to interaction involving Hopewell societies in the Lower Ohio valley represented by Wyandotte chert in this analysis.
It is only as we leave the regional scale and engage with scales of production that exceed the reach of Hopewell societies in the Midcontinent, do we see a significant change in this emphasis on household level production and interaction. Obsidian and Knife River Flint, both coming from distances in excess of 1000 miles, are greatly limited in their distribution within the Scioto Valley, and the activities performed at earthwork sites alone seems to control and direct the procurement of these distant exotics. Again, two different groups appear, one featuring obsidian and the relationships of production it represents, the other featuring Knife River Flint. As with the Middle Scale, time does not seem to be a factor in this distinction. There is some geographic patterning to the distribution of Knife River Flint, which is limited to two sites located close to one another on the east side of the valley, but obsidian is not geographically patterned.
Lithic procurement overall is not structured around ceremonial earthwork sites. Lithic procurement is not dictated by proximity. Lithic procurement, and the relations objectified by the distribution, and, by proxy, exchange of certain raw materials appears to be driven by the household and not any specific spatially constrained community. The materials of the earthworks are indistinguishable from those of the non-earthwork sites (with the exception of the exotic materials discussed above). Earthworks are not central to exchange, and do not structure the interpersonal relations we are interested in ultimately. The relations that determine the structures of these networks are formed at the lowest scale we can measure with these materials. The general lack of distinction between major earthworks, minor earthworks, and non-earthwork sites seems to indicate an inversion of the typical view of what and where Hopewell is. If Hopewell is the pattern of interactions, the Interaction Sphere (Seeman 1979), then it is a milieu of independent household (likely extended families) to household relations giving themselves to networks they built themselves for their own needs. Looking back to Knappett’s framing of the appearance deriving from relations, we can interpret this near identity among site types as implying that the earthworks that emerge from these household ties are a more unified appearance of the relations objectified by the lithic material distributions at non-earthwork sites. In the cooperative construction, over time (in most cases, see Lepper 2010 for an alternative explanation for the Newark Earthworks, a complex outside the Scioto Valley), the earthworks are another objectification of the same relationships. Following Knappett’s presentation of Gell (1999) and Strathern’s (1988) views of exchange relations where the givers ‘are giving away part of themselves’ in the exchange, the mutual giving of labor and
Social interactions that define Hopewell 191 materials to construct the earthworks means these built landscapes are the people who make them and the relationships which brought them together. Those American Indians that constituted Hopewell were literally and figuratively putting themselves into their ceremonial places writing not only their sophisticated understanding of the cosmos into the earth under their feet, but also writing their interpersonal relations into the earth itself.
Several of the patterns are similar to those noted by Nolan et al. (2020) in the purely descriptive analysis of all materials analyzed for the SCHoN project (ceramics, copper, and lithics), and echo the findings of one of the first systematic analyses of the Hopewell Interactions Sphere over a much larger scale than the present analysis (Seeman 1979). Namely, the relations objectified by ‘Interaction Sphere materials’ (Seeman 1979) are not hierarchically structured or constrained by a broader community, but access is mediated at smaller scales. Additional data from ceramic and copper artifacts have yet to be analyzed in a formal SNA context. These future analyses will add depth and texture to our understandings of the structure of Scioto Hopewell social interactions.
Turning to the evaluation of the two major models above (see Figure 8.1), the synaptic model predicts a multi-tiered structure of interaction focused on especially major earthwork sites, while the proximity model predicts independent households focusing their interaction predominately on a local community. While this analysis is preliminary, it does not support the synaptic model proposed by Carr (2005a), and certainly does not indicate any tripartite divisions or a hierarchy of earthworks with the Hopewell site at its peak. Instead, it suggests that Scioto Hopewell is a phenomenon largely structured through the frequent interactions of households across local and regional scales. Only as the scale of production exceeds the reach of Hopewell societies do we see that all households become isolated from the network and the earthwork sites - and the production and interaction they structure - take a prominent role. Even at this scale, there is no hierarchy. The individuals who gained access to these exotic Interaction Sphere materials are not localized. The potential exception is the circumscribed distribution of Knife River flint.
The proximity model emphasizing household-level organization is in closer accord with these results. However, there is no geographic pattern to the lithic networks (again, except the Knife River distribution) as we would expect if these were structured communities, or even peer polities (Pacheco and Dancey 2006). The realization that Scioto Hopewell is a household level phenomenon should change the way we seek to explain its rise and its ‘demise’ (see Dancey 1996). Hopewell emerged from decisions about relations and resources at the household level. It was variously constituted in hamlets and small nucleated settlements, with a rise in the proportion of nucleated settlements through time and a possible shift in the types of places and degree of extra-regional interaction around AD 200 (Burks 2004; Dancey 1992, 1996; Nolan and Howard 2010; Seeman and Dancey 2000).
The Hopewell pattern of independent households negotiating interpersonal relations with distant and diverse neighbors is eventually replaced with focused interaction and alliance with a cluster of households in the larger fortified nucleated villages of the early Late Woodland (ca. AD 500-700; see Seeman and Dancey 2000).
While earthworks may be the stage for Hopewell, earthworks do not make Hopewell; Hopewell starts with households and earthworks emerge from the relations and networks already established between households of the Scioto Valley and the greater Hopewell world.
