Tuesday 13 December 2022

Biasing the Record

Once upon a time, there used to be a "never mind the quality, feel the width" myth spread in British archaeology about the PAS database. Allegedly, by their coverage, British artefact hunters were recovering artefacts from lots of the archaeological record giving a much wider coverage than the targeted investigations of archaeologists. Possibly some archaeologists swallowed this as an article of faith rather than a closely-reasoned argument, I don't know. Anyway, it has long been clear that this is not how it works, the sites targeted by artefact hunters are not random, the material taken from them is not anything like a representative sample, and what then got reported was also influenced by a whole load of subjective factors. Until there was proper documentation of those biases (and Katherine Robbins 2014, 'The Portable Antiquities Scheme a Guide for researchers' is not it), it should be accepted that the PAS "database' is more a case of "rubbish in, Rubbish out", no matter "how many" artefacts it has got in it. Now this from a metal detecting forum near you, just a mouse-click away:

Not all finds are equal ... By liamnolan » Sun Dec 11, 2022 6:50 pm
At a club meeting a few years back, the FLO and fellow archie went around the finds tables and selected ONLY those finds that they felt needed prioritising for recording. Good or bad?
I had mates upset that their prized roman or hammy had not been taken away for further examination. However, I took the view that at least some missing pieces of the historical jigsaw of Norfolk had been recovered. I feel that will be the way forward now if there are limited PAS hours. Plus a rewriting of the guidelines for finders so as to lower expectations.
Should there be a charge made to finders for the PAS work? To increase paid hours?
An unpaid team of knowledgeable detectorists initially sifting through submitted routine finds, doing basic ID and recording, reserving more interesting finds for the FLO? Food for thought! Liam
Indeed, and some of the best minds in British archaeology are simply ignoring the implications, and not challenging the claptrap that "metal detecting is good for archaeology". Randomly selected loose objects are not "some missing pieces of the historical jigsaw of Norfolk" that have been recovered, they are fragments of evidence that has been destroyed by removing them with no documentation from their contexts and associations (including the objects the picky Norfolk FLO did not pick up and take away to enter on the "record").  Here's another one on the same topic:
geoman » Sun Dec 11, 2022 8:38 pm
To reduce the workload the PAS management some years ago reduced the number of items an FLO could take in from each finder and also instructed them to be more selective in what they recorded [Here, PMB].
There are high priorities placed on Treasure items as it is these items which provide the headline grabbing statistics each year to bleet that there has been another increase in treasure finds made by the public - all to keep the politicians and paymasters happy. They dont care too much if it is junk Treasure just that it is captured by the process.
FLO's are based at various museums and archeological department offices, who as local partners provide 10 % funding of the post in cash or kind. Usually it is a cosy desk in the corner of an office with a variety of admin requirements thrown in. The rest of the salary costs are paid by the PAS Central Unit from the funds allocated to the PAS by the British Museum. It is a common joke that the local partners provided 10 % of the costs ,but want 100% of the FLO's time. In some respects that is true and local managers have other priority tasks for their FLO's that are not about recording public finds. Dont forget all the PAS data is passed onto the HER's.
Local managers also have requirements to add information and detail to support the Regional Research Agendas for their area or county. This influences what the FLO's should record as priority items submitted by finders. For example in some counties rich in Roman remains high priority finds may be those of the Anglo Saxon period.
You can see that this will immediately skew any attempts to analyse these "data" in any way that is not "dots-on-a-map". This is biased towards an archaedology that is about loose artefacts in a landscape rather than the taphonomy and character of the sites they come from (in a landscape). A landcscape that is reduced to scattered and selected objects in blank space, rather than as part of a palimpsest. In other words, no "data" at all, especially when seen on the level of the archaeology of a site, a microregion or a landscape.  

And of course... "There are high priorities placed on Treasure items ..." is nonsense. As any law-abiding detectorist must know, they are not picked out by FL:Os, it is the finder's responsibility and legal obligation to report any and all potential Treasure items directly to the Coroner. As the Treasure Act is written, the PAS and the FLO have absolutely nothing to do with it.  



3 comments:

Brian Mattick said...

Luckily finds are always from where the finders say they are so that additional difficulty doesn't arise. Imagine if it did! How would you know how widespread the distortion was? 1%, 50%, you tell me.

Paul Barford said...

Actually, my Christmas break project is actually looking at that with a collector (!)... there are some funny old things in that there database when you look with an enquiring mind...

Brian Mattick said...

PAS does warn researchers about sources of bias but no one can tell anyone the degree, particularly how much find spot falsification there is.

You may be able to deduce something but that's as far as it will go. Certainly not as far as a PAS Annual Report. Those remind me of Plato's cave. Appearance ain't reality, however glossy the paper.

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