A comment on Ward et al.’s ‘Insights into the procurement and distribution of fossiliferous chert artefacts across Southern Australia from the archival record’

Abstract

The presence of artefacts made from fossiliferous chert in the Perth metropolitan area and broader Swan Coastal Plain, extending from Geraldton to Dunsborough (Figure 1), has been an enigma for more than 40 years. The evidence from archaeology is inconsistent with regional geology and, as a consequence, geologist John Glover postulated the existence of offshore sources drowned by rising sea levels (Glover 1975, 1984; Glover and Cockbain 1971). Ward et al. (2019a; see also O’Leary et al. 2017; Ward et al. 2019b) claimed that new geological evidence makes this explanation untenable and therefore propose that long-distance trade from the Eucla area near the South Australian border be reconsidered as an explanation. However, their argument ignores the substantial archaeological evidence that led Glover to his original conclusion, as well as more recent investigations in southwestern Australia. Glover (1975) originally identified two main types of chert in archaeological contexts in the southwest of Western Australia: opaline and chalcedonic chert (referred to as Plantagenet chert), and fossiliferous chert (sometimes referred to as Bryozoan chert in the literature). Other chert types do occur, but in negligeable quantities. Plantagenet chert is common in archaeological assemblages along the south coast from Albany to Esperance, and is clearly derived from local onshore sources of silicified Plantagenet Group rocks. It is recorded in some inland areas, but there is no evidence that this material travelled as far as the west coast. Fossiliferous chert is found in archaeological sites along the west coast. No local sources are known, but it most closely resembles chert from the Eucla area. However, the large quantities of fossiliferous chert found in some sites suggested that it was unlikely to have travelled long distances, and the percentage of fossiliferous chert shows a fall-off from west to east, suggesting a westerly source. Thus, Glover proposed the hypothesis that fossiliferous chert derived from offshore sources drowned by rising sea levels. The effect of distance from source on the composition and characteristics of archaeological stone assemblages under different scenarios of procurement has been well-investigated. Broadly, the representation of stone in assemblages diminishes as distance from source increases, and in the case of highly-valued materials, particularly where highquality materials are scarce, a range of economising behaviours is normally observed. Assemblages along the south coast of Western Australia show precisely this pattern with respect to local Plantagenet chert. A study of the distribution of chert along a transect inland from Bremer Bay showed that the percentage of Plantagenet chert in assemblages diminishes with distance from coastal sources in a characteristic falloff curve. Assemblages inland of Bremer Bay also showed clear evidence of economising behaviour such as reduced size of artefacts and more intensive reduction of cores and tools (Bird 1985). Similarly, the percentage of chert in assemblages in the Albany area increases with proximity to sources in the east of the study area (Ferguson 1985). By contrast, the distribution of fossiliferous chert in assemblages along the Swan Coastal Plain suggests a westerly source (Figure 1). Although Plantagenet chert is recorded in sites at least 150 km inland, it has not been documented in archaeological assemblages on the Swan Coastal Plain. If large quantities of chert were travelling considerably longer distances westwards, from the Eucla area in the east, as claimed by Ward et al. (2019a), it is surprising that closer south coastal sources of Plantagenet chert were apparently bypassed. Ward et al. note that Eucla chert is found in archaeological sites up to 700 km to the east. However, by the time it reached the limit of its easterly distribution, Eucla chert in archaeological sites shows exactly the sort of evidence for conservation that would be expected (Nicholson and Cane 1991).