PAS and controlled archaeological excavation

The Portable Antiquities Scheme (PAS) "allows finds discovered by members of the public to be recorded for the benefit of researchers and the public alike" (Michael Lewis, press release). And I noticed that The Times on Saturday (Mark Bridge, "For history and riches go treasure-seeking", 27 December 2014) was discussing how PAS had reported its millionth find. Yet we also know that the PAS database includes material from "Controlled archaeological excavation". So a major Roman coin hoard excavated by archaeologists in Bath appears in the database.

One of the things that was discussed by Gill and Chippindale is the difference between objects with a secure archaeological context (a1), and those with a reported or alleged find-spot. To what extent is the PAS database falling into the 'a2' (or a3 / a4) categories? How far can we trust reported find-spots? ['a' stands for archaeology. And this is another reason why I am trying to discourage the use of the obsolete term "provenance".]

There are clearly some important methodological issues that need to be addressed in the nuanced design of the database.

And perhaps this is where the Micropasts project enters the debate.

It is important to notice that even quality 'tabloids' like The Times do not check the facts behind press releases.

Micropasts and issues with PAS data

Neil Wilkin on crowd sourcing the British Bronze Age #micropasts pic.twitter.com/K9DUeZAsQo
— David Gill (@davidwjgill) November 6, 2014

It was instructive to listen to Dr Neil Wilkin yesterday at the Society of Museum Archaeology annual conference. He was talking about the Micropasts project and the use of crowd sourcing. It was good to hear a discussion of the digitisation of the card files as well as the images from the Horsfield archive (see here).

At one point Wilkin appeared to have to defend the intellectual reliability of the data provided by PAS. I think that he is right to be cautious. How far can we trust the information supplied with the reported objects? Are these largely reported or "said to be" findspots? Is the PAS information triangulated by more secure information from the Micropasts project?

And what about all the unreported finds?