My recent article offered a model for better classifying feasts and their social consequences by distinguishing between archaeological correlates of group size and sociopolitical competition. Applying this model to Feltus, a precontact Native American mound site, I concluded that feasting events focused on creating and maintaining group solidarity in the absence of significant sociopolitical competition. Given Brian Hayden's broad experience with synthesizing archaeological studies of feasting, I read his comment with interest, and I appreciate the chance to respond. Hayden offers two primary critiques. The first concerns the model itself, and he argues that it does not appropriately deal with certain documented types of eating events. The second questions my noncompetitive interpretation of the Feltus feasts, and he suggests alternate possibilities to explain the recovered assemblage. I address both here.

Hayden proposes two ways my model falls short of contending with ethnographically identified types of feasts. First, he asserts that it does not account for the fact that large-scale, competitive feasts sometimes include small groups that consume special foods in special ways and larger groups that create less remarkable refuse. This critique fundamentally misunderstands how the dual-dimensional model works. Nowhere do I state that only high-status goods would occur at sites that hosted such feasts; rather, I argue that large-scale, competitive events would have created large amounts of ceramic and food debris (i.e., that created by Hayden's larger group), some of which would be distinctive (i.e., that created by Hayden's smaller group). Second, Hayden contends that I do not consider work feasts; however, because work feasts occur in both nonhierarchical and hierarchical societies and have a variety of social impacts (Dietler and Herbich Reference Dietler, Herbich, Dietler and Hayden2001), they can be usefully differentiated by my model. For example, an Amish barn raising (Long Reference Long2003) would plot in the large-scale communal quadrant, whereas feasting associated with corvée labor (Dietler and Herbich Reference Dietler, Herbich, Dietler and Hayden2001:244) would rank higher on the competition scale. Some Feltus feasts occurred during periods of mound construction, whereas others predate it. Consequently, it is likely that the former represent work feasts, albeit ones with little evidence of sociopolitical competition. The data from the various Feltus feasts are broken down by site context and presented in detail elsewhere (Kassabaum Reference Kassabaum, Peres and Deter-Wolf2018).

In critiquing my interpretation of the Feltus data, Hayden raises the possibilities that elite refuse was overwhelmed by nonelite refuse, destroyed by modern disturbance, or not uncovered by our excavations. I do not deny these possibilities, but such uncertainty is pervasive in archaeology. We must interpret the data we have, not those that might hypothetically exist. If we always assume that we are merely missing the evidence that proves our expected conclusion, then we neglect the point of scientific research. Instead, we must weigh the evidence carefully in the context of site condition and the amount and character of excavation. Feltus excavations sampled feasting deposits across the entirety of a well-preserved, well-delimited site, including those associated with mound summits and construction episodes, premound deposits, and off-mound areas (Kassabaum Reference Kassabaum, Peres and Deter-Wolf2018). Hayden also challenges my interpretation of aspects of the ceramic and faunal assemblages. Hayden's suggestion that all pottery at Coles Creek sites was for specialty food preparation and competitive display is unreasonable. The citations he offers in support of this assertion pertain to the earliest iterations of ceramic technology, but pottery had been in widespread use in the Lower Mississippi Valley for over 1,500 years before the Feltus feasts (Phillips Reference Phillips1970). Similarly, Hayden contends that bear ritual and pipe smoking were used to create inequality rather than social cohesion, going so far as to suggest that human sacrifice explains the Feltus ritual deposits. In doing so, he cites data from Siberia, Japan, the Northwest Coast, the Great Lakes, the Plains, and the Southwest, but never the Southeast. My own thorough reviews of bear (and pipe) symbolism among southeastern groups (Kassabaum and Nelson Reference Kassabaum and Nelson2016; Peles and Kassabaum Reference Peles, Kassabaum, Lapham and Waselkov2020) simply do not support his conclusions. Moreover, the presence of human bone in ritual deposits is to be expected on Coles Creek sites, where bone handling and secondary burial was prevalent (Kassabaum Reference Kassabaum2011). The differences between Hayden's interpretation of the nonhistorically or geographically specific ceramic and faunal data and my interpretations of the comparative data drawn from relevant southeastern sources clearly emphasize the importance of interpreting feasting events within their cultural and historical contexts.

In conclusion, I would welcome Hayden's equivocation about the nature of feasting at Feltus if it were offered alongside geographically and chronologically appropriate data, but such information is largely not provided. Instead, I believe Hayden's critique stems from fundamental differences in our assumptions about the nature of humans and society. The schism between those who believe humans are inherently competitive and those who are focused on cooperation is well documented (e.g., Pluckhahn Reference Pluckhahn and Carballo2013:177) and certainly at play here. For example, Hayden states unequivocally that group solidarity is only important under competitive pressure and that mounds were constructed as displays of power, but the data he offers in support of these assumptions are not contextually relevant. Moreover, by taking these assumptions as his starting place, Hayden places the entire burden of proof upon those invoking group cohesion and none on those invoking competition. This is akin to placing the onus on gender scholars to prove something was done by a woman while allowing one to just assume it was done by a man. I see my broader work and this article as joining a range of scholarly voices that are explicitly challenging these assumptions and encouraging conversation among those who disagree about such fundamental issues (e.g., Carballo Reference Carballo2013). The types of questions posed by Hayden in his comment—such as “who [or what] promoted the construction of mounds and the gathering of dispersed homesteads, and why was this important?”—represent precisely the types of questions we should be asking. We just need to be most open to answers supported by historically and culturally appropriate data.