By Rachael Sparks, on 19 March 2012
There is a legend that when every object in a collection has been given a unique accession number, its curators will be freed of the shackles of performance indicators and documentation plans and finally achieve a state of nirvana. There’s lots of self-help guidance out there, of course (deep breathing exercises optional) to help us achieve this goal, including information on how and when to number objects. The sensible way, according to the Collections Link’s subject factsheet, is to give objects a running number, or, if you must, a number representing the accession year and then a running number. So surely that’s what everybody does, right? Wrong!
The reality behind the ideal is that many collections were assembled a long time ago when standards were – how shall I put it? – somewhat looser than they are today. Curators have therefore inherited all sorts of interesting, not to mention inpenetrable, inconsistent and sometimes downright unworkable systems. All of which are usually exacerbated by the fact that these older systems were rarely documented properly. So even figuring out how they are supposed to work can be a challenge in itself.
The Institute of Archaeology Collections at UCL are no exception.
There was some kind of accessioning system in place even before the Institute opened its doors for business in 1937. Whoever designed this system began it in a flurry of inventiveness. Lets make accession numbers useful. Let them record more than the cold, hard fact of the individual nature of an object. Let them speak, through their form, and tell us much more of their personal stories. From such idealistic beginnings, came creative encoding of all sorts of data into each accession number.
So our first accession numbers begin with a nod to the geographic part of the world in which they were found. Actually, number is a bit of a misleading word here, as the first thing you get to read is a letter. The system was conceived as being truly global, so letters were assigned even to regions that weren’t represented in the collections – yet. Perhaps this speaks to the Institute’s ambitions – or pretensions. So we have E for Palestine, F for Syria, G for Mesopotamia and so on.
So far so good.
The next part of the number was designed as a code to represent the site. Very much in keeping with the tastes of the time, this was presented in the form of Roman numerals. Simple it wasn’t, but oh so very elegant. So now we can move to numbers in the form of EXI for Tell el-‘Ajjul, EXIX for Jericho, EXXXVI for Tell Jemmeh and so on.
Of course, this got everybody thinking. If we can code in geography, then why not chronology? Why not also divide the numbers up to represent different periods at that site? Following this rule, this allows us to break up material from Tell el- ‘Ajjul, for example, into EXI (Early Bronze Age), EXII (Middle Bronze Age) and EXIII (Late Bronze).
A slight inconvenience of this system – not obvious to its designer – was that you have to be able to date an object before you can give it an accession number. Ever tried to date an unstratified burnishing pebble? Not something for the faint hearted.
I guess others came to the same conclusion, because this system only lasted a short time. After working through material from a handful of sites it was eventually abandoned in favour of the more pedestrian but practical one site = one code approach.
But wait; we’re not finished with our numbers yet! Two more parts to go.
The first of these is a single number representing a group of material within the designated site. On some sites, this meant everything from a single context, like a tomb. On others, to be honest, its not too clear what it meant. This was then followed by – at last – a straightforward running number.
Put it all together, and we might get something like EXII.33/1, representing item 1 from tomb 429, an Middle Bronze Age context from Tell el-‘Ajjul.
Are you still with me?
This beautiful coding does mean that I can pick up one of these old objects, read the number and four times out of five recognize the site it comes from. But sometimes our curatorial staff just got carried away in a riotous romp through the many possible permutations of letters, Arabic and Roman numerals.
Take the tragic story of basalt bowl fragment EIII.1iig/3.2.
Yes that’s right; a seven-part number this time. Not only lengthy (try writing that on a bead!), but also crying out to be misread and misunderstood.
So here we have The letter ‘E’, followed by Roman numeral ‘III’, then Arabic numeral ‘1 that looks frustratingly similar to the preceding Roman numeral, while being poorly distinguished from the following lower case Roman numeral ‘ii’ – after which we have a ‘g’, that looks very similar to the way some people write the number ‘9’, followed by a forward slash, that when poorly written is easily mistaken for a ‘I’ or a ‘1’. The last two numbers are probably fine, providing the ‘3’ is written clearly and doesn’t accidentally morph into an ‘8’, and the base of the 2 doesn’t wear off and make it look something like a ‘7’.
But now we are moving into the dark territory of differing handwriting styles, another thing that is sent to try us and which can no doubt be saved for another confessional.
After all this, if researchers then go on to record and publish our accession numbers incorrectly, how can we possibly complain?
These long numbers took their toll, and by 1937 (the ‘official’ opening of the Institute) a decision was made to abandon the whole system and replace it with – you guessed it – a simpler year numbering system. Even then, there was a period of transition in which the two systems overlapped, showing a reluctance to let the idea of coded data go completely. There was a phase in which somebody decided to go around and reaccession things with numbers in the new system (see below for why this is a bad idea); and another phase in which they couldn’t quite let go of the area letters, and so introduced a whole new series of letter codes, this time tied into the different teaching departments (somewhat oddly as F for Western Asia, Z for European, and S for Indian). But the dust settled, and as Britain entered into a bold phase of post-war optimism and modernity, the whole concept of letters and Roman numerals were for the large part left by the wayside.
So should we all tackle our obsolete numbering systems and start passionately renumbering everything? Hell no! That’s the worst thing you could do, according to Collections Link advice. If it works as a unique identifier, then however clumsy it might look, it still works and there’s no need to fix it. The quaint, elegantly awkward systems of the past are all part of the rich heritage of our profession. What we need to do is stop fighting, embrace the peculiarity of it all, and pass on this knowledge with relish to every unsuspecting researcher or student who comes our way. If only to see the incredulity/amusement/despair on their faces.
Ah well, nirvana schmana. I expect enlightenment is a bit like retirement; sounds great, but then you don’t know what to do with yourself afterwards.