Reply to McPherron et al.: Doubting Dikika is about data, not paradigms

We critiqued the claim that bone surface marks on two ∼3.4 Ma fossils from Dikika (Ethiopia) are the earliest evidence of hominin butchery damage (1) by (i) providing a detailed argument showing site- and assemblage-level weaknesses for the claim and (ii) matching those marks morphologically to marks produced on modern bones by trampling in coarse-grained sedimentary substrates, similar to those of the Dikika site (2). The letter by McPherron et al. (3), responding to our critique, completely ignores these substantive issues and, instead, seeks to disqualify our assessment by arguing that we are unwilling to accept a paradigm shift. McPherron et al. (3) still provide no compelling evidence to support their claims that the Dikika marks are butchery damage, and their posthoc rationalizations (3) require response.

• i) Their assertion (3) that the stratigraphic provenience of the fossils is identified is not sustained. The Dikika fossils are ex situ finds. Despite considerable justification of their inferred provenience of the fossils (1), McPherron et al. (3) cannot show the precise lithological context (e.g., clay, silt, sand, or gravel) in which the fossils were originally deposited. They suggest that the fossils may have come from a sandy deposit (1). Sands possess abrasive properties. Thus, trampling (and/or other incidental movement on or within the substrate) of bones should not have been ruled out by McPherron et al. (1, 3) as a significant potential taphonomic process operating at Dikika.

• ii) McPherron et al. (3) are also unsuccessful in their attempt to invalidate the stratigraphic control of evidence for 2.5–2.0 Ma butchery from Bouri and Gona (Ethiopia). The stratigraphic contexts of cut-marked bones from these sites were, in fact, established through excavations of in situ fossils (4, 5). Furthermore, far from deriving from “similar or inferior” (3) contexts to Dikika, the Bouri and Gona samples are instead from silts and clays, sedimentary contexts that make trampling damage on bone surfaces unlikely.

• iii) The claim that our bone trampling experiments did not produce a single bone-surface mark that “remotely resembles” (3) the deep, V-shaped Dikika marks, DIK-55–2-A1 and -A2, is simply false. We illustrate such trample marks in figure 4 A and F in ref. 2.

• iv) The statement (1) that the Dikika fossils do not display surficial microabrasion, typical of trampling damage, is false (2). The claim (1) that striae fields on bone surfaces result solely from the application of stone tools is incorrect; striae fields can also result from trampling (2).

Given the magnitude of the interpretation of Dikika forwarded by McPherron et al. (3), a conservative (configurational) approach to the interpretations of these marks is epistemologically mandatory (2). The null hypothesis required here is that the Dikika marks result from natural abrasion. By matching the Dikika high-confidence “cut marks” with essentially identical, known experimental trample marks, we showed that the null hypothesis cannot be falsified with existing fossil evidence from Dikika. Our dubiousness about the Dikika marks as butchery damage does not emanate from resistance to paradigm change but simply from the lack of any scientific support for that remarkable claim.