An austere realm of great beauty

Being an archivist is sometimes a strange profession, spanning a range of worlds. Your training can include instruction in Tudor handwriting or medieval Latin, but will also cover nuts and bolts information about reading room layouts, order slips and avoiding damp in your strongroom; whilst the working day can take you to discuss one of our medieval treasures with a scholar planning a critical edition, or into a dark garage or basement to survey papers covered with decades’ worth of cobwebs. It is surprising how often the ability to drive a Transit van through a narrow gap comes in handy, too: and an archivist in their first job soon learns that perhaps the most crucial people in their entire building are the people who run the loading-bay. We cover a range, then, from the elevated and academic to the sternly down-to-earth. By and large, though, archive work is practical: it centres on the basic task of acquiring material and carrying out the range of jobs required to get a description of it onto the catalogue and the thing itself into the reading room (or, increasingly, onto a screen as a digital facsimile).

Of course we theorise: we discuss what to acquire and (just as significantly) what to leave behind, and we agonise over whether our acquisition policy and catalogue descriptions can ever be neutral and free of bias. Just to prove that archive work is not theory-free, we can point to the fact that Derrida has written on the subject (Archive Fever): writings don’t get much more theoretical than that. But our basic principles are rooted in pragmatism. The most fundamental principle we have, that in cataloguing material one seeks to reflect the arrangement of the material given to it by its creator (and thus to reflect how they thought about the world, how they categorised things, not how we do those things) grew out of practical experience: the discovery that if you did anything else, if you arranged the papers in a way that reflected some abstract categorisation, this layout grew more and more out of date as the years passed and it became simply impossible to find anything.

It probably isn’t surprising, then, that for some years archivists were suspicious of cataloguing standards: of cataloguing material in accordance with some abstract principle, people feared, rather than having the description grow organically out of the individual item one had on one’s desk. The computer changed that: a database requires one to think about the things that one says about archives and to think about common elements, and the advantages of being able to exchange and pool data with other repositories are sufficiently great for people to start thinking about how one might map one office’s descriptive fields to another’s. It’s not a coincidence that the first widely-accepted international archive catalogue standard came out in the late 1990s, as the internet began to change the world of information: the snappily-named ISAD(G), or General International Standard on Archival Description. (Not long after this we in the Wellcome Library began loading our old archive and manuscript catalogues to a database built around this standard, a process described here.)

ISAD(G) grew out of applied common-sense; a fusion of best practice as it was seen in archive catalogues around the world. Perhaps surprisingly, perhaps not, these various catalogues all converged on some basic concepts. Particularly relevant to those of us designing databases, the profession decided that there were basically twenty-six things one could say about an archive item, twenty-six concepts that would map to fields in a database. Some of these were specific, some very general (and notably number twenty-six is a general “Notes” field that translates to “Anything relevant that you couldn’t fit into the previous twenty-five”): and of them all, only five are crucial. To conform to the standard, a record must have these crucial five fields; the others can be left blank.

Earlier this summer I was talking at a conference of postgraduates specialising in the early modern period, explaining to them why they, as users of archive material, saw the things that they did when they consulted an archive catalogue, and what our reasoning was. This involved touching briefly on ISAD(G) and in particular on the five obligatory concepts that have to be there in any description of archive material. These are:

1. Reference: that is, some unique identifier that establishes which bundle of papers, out of all the millions in the world, you mean; for example, a manuscript number such as MS.49. Behind the scenes in our database these references have two little prefixes, identifying archive material as belonging to this library and no other, in this country and no other: this full reference, GB 0120 MS.49, distinguishes our MS.49 from any others in the world that may carry the same number. (GB is self-explanatory; 0120 takes you to our entry in the National Archives’ Archon directory of archive repositories in the UK.)
2. Title; or, if the item doesn’t have a formal title, just a brief description of what the item is (“Correspondence file, A-D” for instance).
3. Date: when was it created? (Even if it may not have an obvious exact date, one can usually at least guess to the nearest century: “1840s-1860s”, “early 20th century” or “16th century – 17th century” would all be acceptable.)
4. Extent: how much of it is there? Essentially, what is the reader up against – are we describing a single sheet of paper, 300 boxes, or some point between them?
5. Level: this is the closest we get to archive technicality: essentially, this asks one to say whether what’s being described is an entire collection, the whole archive of a given person or organisation, or simply part of a larger whole.

Aristotle: a bust found in the ruins of Herculaneum. Image copyright the Clarendon Press, made available under Creative Commons by-nc 2.0 UK: see Wellcome Images, image M0005620

At the end, a philosophy specialist approached me, excited by what he’d just heard: the way that the immense variety of all possible archive material was boiled down to these basic concepts reminded him, he thought, of what Bertrand Russell said about mathematics (in Mysticism and Logic), as having “not only truth, but supreme beauty – a beauty cold and austere, like that of sculpture.” Furthermore, the five ISAD(G) compulsory fields were, he thought, essentially basic logical categories set out by Aristotle in the 4th century BC. I had to take his word for this but on return to the office, a quick look at Wikipedia showed that he was, indeed, onto something. Aristotle sets out ten basic categories, more than our compulsory five, which gives us the opportunity to browse them in search of matches, and sure enough we can set up a tentative mapping as follows:

1. Reference = Aristotle’s “Substance”: that is, the fact of this thing being uniquely one thing rather than another: the equivalent of the way you are yourself and not any of the billions of other people who are living or have lived;
2. Title = Aristotle’s “Qualification” (what is it like?), describing it in terms not of its uniqueness but of qualities shared with other similar things (this is probably the most tentative of the mappings, as the sort of descriptive information found in this field could be shared across a few of Aristotle’s ten);
3. Date = Aristotle’s “When”: its position relative to other events in a sequence;
4. Extent = Aristotle’s “Quantity”: how much space does something occupy; and
5. Level = Aristotle’s “Relation”: is it part of something else, or does it encompass that other thing?

Archivist and transit van. Other van hire firms are available.

Aristotle’s reputation in philosophy has long been one of pragmatism (contrasted, for example, with the “idealistic” Plato). It’s fitting, then, that a cataloguing standard based on pragmatism, on distilled best practice rather than advance theorising, should end up mirroring his work. We’ll avoid getting into any discussions about “reality” at this point, and whether this indicates that this is, therefore, how the universe “really” works – this is as far into philosophy as I want to go – but archive cataloguers can get a little glow of satisfaction from this. Readers using our catalogues can pause for a moment and look around at the austere realm of great beauty in which they find themselves. And Aristotle, the practical down-to-earth philosopher, can be inducted into the profession as an honorary archivist; although I’m not sure how good he’d be at driving a Transit.

[Edit: this post describes the first edition of ISAD(G), to which Wellcome archive catalogues conform.  Subsequently, the second edition added a sixth compulsory field, Creator Name, which is not yet used on all Wellcome archive records.  There’s probably an entire blog post to be got out of how one defines the “creator” of a large archive: the person who writes most of the actual documents, the authors of letters to the person whose papers these are, the person who gathers all the documents together….]