12th June 2011, 03:02 PM
Unit is right. no single view should be taken as gospel. Just because it has Bayesian in the title does not mean all other epistemological issues have been dealt with. Archaeological data is very fuzzy data.
However, as a tool, Bayesian statistical approaches relying on mutually connected probabilities of differing events is as good as any other - used badly it is like a shovel to pick teeth.
but it is an excellent tool for trying to untangle webs of connected probabilities....whether Bayliss et al have got this one right, will require review from persons with similar expertise. The results should not however predetermine approaches to archaeological sites or the material evidence they contain. Other issues (taphonomy, individual site integrity etc) might also be considered in detailed critique of particular sites.
However, as a tool, Bayesian statistical approaches relying on mutually connected probabilities of differing events is as good as any other - used badly it is like a shovel to pick teeth.
but it is an excellent tool for trying to untangle webs of connected probabilities....whether Bayliss et al have got this one right, will require review from persons with similar expertise. The results should not however predetermine approaches to archaeological sites or the material evidence they contain. Other issues (taphonomy, individual site integrity etc) might also be considered in detailed critique of particular sites.