Thursday 6 April 2017

More UK Archaeologists Approach PAS Database Uncritically


Cooper, A. and Green, C. 2017 Big Questions for Large, Complex Datasets: approaching time and space using composite object assemblages, Internet Archaeology 45.
This study tackles fundamental archaeological questions using large, complex digital datasets, building on recent discussions about how to deal with archaeology's emerging 'data deluge' (Bevan 2015). At a broad level, it draws on the unprecedented volume of legacy data gathered from many different sources - almost one million records in total - for the English Landscape and Identities project (Oxford, UK). More specifically, the paper focuses in detail on artefact evidence - material derived primarily from surface surveys, stray finds and metal detecting. 
and guess where they got that from? Yep this is another PAS-inspired fluff piece.
Novel computational models are developed that extend and connect ideas from usually distinct research realms (different arenas of artefact research, digital archaeology, etc.). Major interpretative issues are addressed including how to approach background factors that shape the archaeological record, and how to understand spatial and temporal patterning at various scales. Overall, we suggest, interpreting large complex datasets sparks different ways of working, and raises new theoretical concerns.
Except they do not discuss what the difference is between what collectors do to archaeological sites versus what others do... Quite a fundamental point one might think. I am totally unconvinced by their manner of calculating the statistics of the percentages of known sites being targeted as potentially productive sites by these collectors. My suspicion, based on knowledge of how artefact hunters go about 'researching' a region is that there is methodological error in the way they are counting this. Something to look into.

No comments:

 
Creative Commons License
Ten utwór jest dostępny na licencji Creative Commons Uznanie autorstwa-Bez utworów zależnych 3.0 Unported.