Zeder: "The Origins of Agriculture in the Near East" (2011)

Traditionally, archaeologists have taken the hallmark of domestication to be genetic changes: A plant or animal counted as domesticated when it differs from the wild land species it was cultivated from.

In this interesting review paper, Melinda A. Zeder of the Smithsonian Museum mentions a number of reasons that the genetic criterion might mislead us into adopting dating domestication unduly late:

• Ancient threshing methods might have involved beating cereal straws into a basket to get the grains out, a practice that would disproportionately select grains from weak-stemmed straws. This would counteract the effect of crop selection, where the plants are usually selected to have stronger stems. (p. S226)
• When harvests failed, as must have done once in a while, the crop may have been restocked from the wild land species, wiping out the genetic record of domestication. (p. S226)
• Occasionally, the morphological changes we are looking for are in fact quite modern phenomena. Zeder mentions that the large size of modern broad beans, a trait that did not appear before well into the middle ages. (p. S225)

In addition to these rather technical reasons, a more important issue is that a too rigid criterion can lead us to ignore or misunderstand practices somewhere on the continuum between opportunistic gathering and habitual farming. Such hybrid relationships between land and people, Zadeh writes,

… not only makes it impossible to identify any threshold moments when wild became domestic or hunting and gathering became agriculture but also shows that drawing such distinctions actually impedes rather than improves our understanding of this process. Instead of continuing to try to pigeonhole these concepts into tidy definitional categories, a more productive approach would be to embrace the ambiguity of this middle ground and continue to develop tools that allow us to watch unfolding developments within this neither-nor territory. (p. S231)

I second that.