> what you did sounds wonderful, but only practical in an academic
> context where you have someone paid to enter the data into the
> database. If registering a find became even vaguely like filling out a
> tax form, then not only would people avoid registering artefacts, not
> only would the database loose credibility, but more than likely the
> information would be so complex that it would almost impossible to
> retrieve in a meaningful way.
Paid? --- I'm not familiar with that term ;-)
If entering something became like filling out a tax form _I_ would not
enter anything! I have two years of tax forms to fill out at the moment
and have a phobia about such things, even though I know I will be
getting some money back. Unfortunately, I _have_ to do it really soon.
> Let me suggest various levels of database entry:
> 1. Self-entry by the public.
> The data would be that which could be obtained without expert
> knowledge - place, pictures, and not much more. The great advantage is
> that after the database is set up, the manpower to run the system is
> very low.
> 2. Additional data from public self-entry
> In order to categorise information such as type of coin in a
> meaningful retrievable way, the self-entry data has to be analysed by
> a relevant expert. Since I would include all archaeological artefacts,
> the range of experts required would be quite huge (and often the task
> would be difficult with just a photo and require follow up). I don't
> know what kind of volumes we are talking about, but I doubt this would
> be a trivial job even to give a name to all the artefacts (if the
> database were successful .... along with all the good artefacts will
> come all the Victorian terracotta garden pots.)
These levels might be bumped up a bit as far as details by the addition
of a wizard, or better still, an expert system at the front end. Also,
perceptual engineering could be employed in a number of ways: the
selection of the right colour combinations to create calm and interest
and aesthetically pleasing interfaces with variety built into subsequent
screens to increase interest, curiosity about what follows and to
eliminate boredom. Additionally, the GUI could take on the aspect of a
game with rewards of "levels" assigned to the user. Animations, avatars,
could also be included.
One of the most brilliant applications of this way of thinking were the
U.S. Army training manuals designed as comic books! Today, the
traditional comic book has entered the world of fine art with Roy
Lichtenstein and the current peoples art is the graphic novel with its
increasingly high production values that derive from computer games and
the computer game itself -- the latter even influencing film.
For your remaining levels, which are valid concerns, their success
depends first of all on what their purposes would be. The PAS is next to
useless as a research tool: it does not contain archaeologically
excavated items, museum objects, contents of private collections and
records culled from the trade. The Celtic Coin Index did contain all of
those things. I wish that I could take credit for that, but the practice
of doing so was started back in the sixties. The PAS version has not
made that particularly visible and this was a political decision. What
people do not understand is that the PAS is really a cult. It is not
intended to be a workable tool and that design flaw starts from the
beginning because it is used only for thing found by members of the
public. It's one big "sampling error". Its interface is geek-like and
"right out of the box" -- why on earth would anyone want a "surprise me"
query?!!! It's cult status can be favourably compared to the story of
the Emperor's new clothes.
The only sensible approach is to start with classes of objects (as was
done with Celtic Coins (Oxford) and Early Saxon Coins (Fitzwilliam). I
introduced "associated finds", "attributes" and "subsidiary elements" to
the on line CCI, but nobody got it and there was never any suggestion to
fill these fields with data. They were put in there as part of our
"Arethusa" database concept which is no longer being worked on. This
would have ended up, in the distant future as a linked series of
collections and museum databases with a web interface all operating
under the same system. The World Bank thought it was great plan. It's
application would not have been limited to cultural concerns and could
also have been used in other sciences and humanities.
The world is really not ready for such things -- perhaps in a century or
With respect to classes of objects, and specifically to early Celtic
art, the current literary version is to group objects by type
(Jacobsthal, Jope, and most likely the Megaws (forthcoming). The
advantage in a database medium is that design elements can be separated
out and queried as such -- which is very important as they span
different types of objects. It is a nightmare to try to do this sort of
thing with a book Of course, the same can be done with material analyzes.