John Marston published a new article in Current Anthropology 65

 

Excerpt from https://www.journals.uchicago.edu/doi/full/10.1086/731785#_i27

Agricultural Entanglements, Niche Construction, and Low-Level Food Production

Archaeology Program, Boston University, 675 Commonwealth Avenue, Boston, Massachusetts 02215, USA (). 22 I 24

Zhuang and Fuller document the intertwined domestication histories of millets and pigs in North China, making four connected arguments: the process of domestication is slow and extended, the combination of theories of entanglement and niche construction is productive in explaining domestication, domestication results in multiple trajectories for subsistence economies, and ecotopes (the smallest spatial unit of ecological function) that sit along ecotones (boundaries between larger ecological units) are particularly productive locations to study for the origins of domestication. None of these, individually, are a new insight. It has been well established based on both genetic (Allaby, Ware, and Kistler 2019) and archaeological (Fuller, Asouti, and Purugganan 2012; Purugganan and Fuller 2011) evidence that processes of initial domestication of both plants and animals are slow and protracted, regardless of taxon or world area, including specific evidence for broomcorn millet (Stevens et al. 2021). The connection of niche construction with theories of entanglement to explain both initial domestication (Fuller and Stevens 2017) and agricultural development (Eriksson and Arnell 2016; Marston 2021) has offered insights into the processes leading to increasing reliance on domesticates. Multiple economic trajectories resulting from domestication have similarly been acknowledged for decades, with “incipient agricultural” (Steward 1949) economies and societies engaging in “low-level food production” (Smith 2001), two terminologies for the “middle ground” between hunting-gathering and agriculture. The unique valuable environmental dynamics of ecotopes located at ecotones have been documented ecologically (Troll 1950), culturally (Turner, Deur, and Mellott 2011), and archaeologically (Spengler, Frachetti, and Fritz 2013).

If none of these arguments are new, then what are the original contributions of this paper? Zhuang and Fuller unite these four key arguments to illustrate, for the first time, the full complexity and diversity of millet and pig domestication in North China. One particular strength of their argument is the depth of ecological study that underlies their geographic analysis of sites of early domestication. Rather than focusing only on macroecology, such as the ecotones between forest and steppe, the authors incorporate detailed study of climate and soil character in locating particular ecotopes that supported millets uniquely well. Attention to soils, especially the dynamics of loess deposits in particular locations (e.g., hillslopes vs. river plains), helps to establish the affordances of specific ecotopes for human interactions with plants that may have led to domestication relationships. Such attention to soils is rare in domestication studies but explains agricultural dynamics in other regions, such as in India (Bauer 2023; Bauer and Kosiba 2016), where the affordances of soils in their responses to human action set in motion a chain of economic relationships that became regionally culturally entrenched. Such analysis helps us to focus, as Zhuang and Fuller argue, on landscape as the object of study rather than centering on human actions or plant or animal biology alone, as in prior approaches. The advantage of a landscape approach is that it permits a more comprehensive understanding of the factors that led to the diversity of early agricultural practices seen in the archaeological record. Together, these factors help us to understand why early millet domestication occurred where it did, an important contribution.

This study does leave additional questions unanswered. Perhaps chief among these is why domestication of millets and pigs occurred when it did. Earlier scholarship linking climatic change at the end of the Pleistocene and beginning of the Holocene to the domestication of plants offers incomplete explanations for the timing of this phenomenon, even when incorporating cultural processes (e.g., Richerson, Boyd, and Bettinger 2001). Zhuang and Fuller argue that reductions in natural fire during the mid-Holocene led to the need for human-mediated disturbance (via either fire or pig rooting) to maintain millet-rich ecotopes, which might explain the timing in both additional investment in millet management and increasing control of pig populations. This does not, however, explain the timing of the ultimate origins of millet domestication, which requires additional research.

Another question yet unanswered is what happens when initial domestication does not result in the development of an agricultural economy. Zhuang and Fuller usefully observe that such cases appear in their datasets but view these as “dead ends” that failed to persist “for long enough to lead to entrenched predomestication cultivation and ultimately domestication,” such as in the case of Houli. Such a perspective treats agriculture as an end state, but it might be more useful to consider that low-level food production making limited use of domesticates may in itself be a desirable end state rather than a failure or “dead end” on the way to agriculture (Marston 2017; Smith 2001). Further attention to the question of how such alternative outcomes of domestication arise is warranted.

To do so, I suggest that alongside concepts of entanglement and niche construction, this discussion would benefit from the integration of broader concepts of ecological and cultural inheritance, which can prove advantageous in considering how and why certain agricultural trajectories arose. Legacy effects, the “persistent impact of ecological interactions” on a landscape that structure its current form (Cuddington 2011:203), are of particular value here, as they are more broadly in agricultural studies (Marston 2021). We can find legacy effects in this case study in growing human dependence on mosaics and ecotones that require fire or animal disturbance, land clearance of food-producing trees resulting in cultivated cereals shouldering a greater caloric burden, and shifting soil hydrology that favors specific plants over others. Similarly, it would be useful to conceptualize tipping points or thresholds (ecological or cultural) that lead to one outcome (e.g., agriculture) versus another (e.g., low-level food production) from an original landscape in which many agricultural opportunities are available (Marston 2015; Walker and Meyers 2004). Parallel structures of agricultural dependence may arise from such tipping points, which, if identified archaeologically, might help us to understand better why certain elements of domestication occurred when they did: for example, why did seed size change or animal size change when it did and not earlier? I suggest that such consideration will give us greater insights into the complexity of domestication and early food production.