I’m at the British Library for the next couple of days for the JISC/BL Discovery Summit. This is an event that brings together work from the last 4 years which started with the snappily named “Resource Discovery Task Force” which was asked to consider how ‘resource discovery’ (finding resources in libraries, archives, museums etc.) could be improved for UK HE researchers and others.
I’m facilitating various sessions at the event, but I’ll try to blog some sessions as well. But I thought I would start by publishing a presentation I did for the first meeting of the Resource Discovery Task Force back in November 2008. I was asked to consider “What if we were starting from scratch”. This was written 5 years ago, and I’ve left the text as it was – so changes in direction and thought in the last five years are not part of this presentation – but I think I still believe in the fundamentals of what I said here, and I think the last 5 years work on ‘Discovery’ have borne this out.
What if we were starting from scratch
What if we got rid of services, standards, methods and mechanisms that are currently make up our resource discovery infrastructure? In 10 minutes!
How far back do we go? To the very first ‘library catalogues’ – written in Cuniform? Or possibly just as far the Library at Alexandria (3rd century BC). In a comment on my blog Joy Palmer (MIMAS) said “how far do we get to go back for a clean slate here? couple of hundred years? before the emergence of bureacratic cataloguing and classification practices? probably not;-)”
Well, I’m not going to go back that far. In fact, I’m going to start by going back just as far as 1931 when Shiyali Ramamitra Ranganathan published his ‘5 Laws of Library Science’
In a recent blog post Lorcan Dempsey reflects that although there have been numerous attempts to ‘update’ the 5 laws they are not particularly convincing, and that this is perhaps because that the laws as they are continue to capture the fundamental challenges of what we do.
All 5 laws are worth exploration, and several of them touch on the intimate relation between resource discovery and access to those resources which is an aspect of the infrastructure I believe is vital, but I’m not going to consider in this presentation. Focusing solely on the ‘Resource Discovery Infrastructure’, I think that the last two are particularly relevant.
The 4th Law – Save the Time of the User.
Ranganathan was concerned with shelf arrangements, recent acquisitions shelves, signposting and shelf labelling. In one example he notes that the activity “may look like a great expense when considered from the isolated library-point-of-view, … can be seen to be really economical from the larger community-point-of-view” He also related saving time of the user to saving the time of the staff – what is efficient for the user is efficient for the library.
One of the major time sinks for a user in our current Resource Discovery environment is the number of places you have to look for stuff.
The 5th Law – A Library is a growing organism – or perhaps because we can consider in this context there is more than one library – libraries are growing organisms. This the law that Lorcan Dempsey was specifically commented on in his blog post, considering how we need to think about effective service in a networked environment. Ranganathan commented “an organisation which may be suitable for a small library may completely fail when the library grows big” Ranganathan asked of the library catalogue “let us examine what form the Fifth Law would recommend for it”. He suggests that libraries pioneered the use of loose-leaf binding to update catalogues, and then the use of cards – with one entry per card – describing the card index as “another epoch making contribution of the library profession to the business world in general” I believe that these principals outlined by Ranganathan can inform how we should design a Resource Discovery Infrastructure that servers the resource explorers. Before I come back to the question of ‘starting from scratch’, I think it may be useful to just reflect briefly on how we have got to where we are today.
I want to briefly go back further to the end of the 19th Century, when Melvil Dewey was involved in the introduction of a standardised form of catalogue card, and catalogue card cabinet (the fact that Dewey also setup the company that sold the cards …
… and the cabinets …
… and even special machines to type the cards may have had an influence in this!
The Library of Congress started to publish it’s catalogue records on these standard sized cards and by this method could distribute them to other libraries.
This was so successful, that Charles Cutter, who produced a seminal work on building a printed dictionary catalogue quickly had to revise it to take account of the card catalogue. By the time the 4th edition of Cutter’s work was published, he prefaced it by saying “any new library would be very foolish not to make its catalog mainly of them [LoC cards]”
This is, I believe, the start of our current Resource Discovery infrastructure – and we have been stuck in this mold ever since. Although we now have computerised catalogues we have never got away from the idea that these records are physical objects – discrete things that are copied for local use. It is for this reason, that when we try to follow Ranganathan’s 4th law to ‘save the time of the reader’ we try to put these copies back together – and encounter problems of inconsistency and duplication. So what do we need to do differently?
One tempting scenario is to think that we should stop ‘copying the cards’, and have a single ‘catalogue’ – it wouldn’t be trying to bring together disparate records – it would be the only copy. When I posted the question on my blog “What if we started from scratch”, one respondent said: “would there be a ‘master’ UK catalogue of books … The trad library catalogue would just be a ‘view’ of this”.
However, I do not believe this is realistic –we don’t have this kind of control over resources and resource discovery systems – we would always have things springing up outside the ‘big store’. Essentially this approach ignores Ranganathan’s 5th law – the library is a growing organism.
To move beyond the card catalogue, we need to look to a more recent figure – Tim Berners-Lee. I believe that if Ranganathan were to look at the problem now, he would recognise that HTML and the web represented that next step that Resource Discovery Infrastructure needed to take to go beyond the card catalogue – but for some reason librarians and archivists have not been in the forefront in the adoption of this as an approach to Resource Discovery, but have treated merely as a new medium with which they need to integrate their existing infrastructure.
When Tim Berners-Lee first described the requirement for a ‘global hypertext system’, he said “Information systems start small and grow. They also start isolated and then merge. A new system must allow existing systems to be linked together without requiring any central control or coordination.” – I think the first part of this statement is a restating of Ranganathan’s 5th law, for a networked age.
I would contend that the Web is the most effective system for disseminating information yet conceived (we might consider the brain an even more effective networked information system?).
What does all this tell us about starting from scratch?
We need to build a linked environment.
The way I think of it is each catalogue record should be a hypertext document. This can both link, and be linked to. When a library adds an item to its catalogue, it can do this by creating a new hypertext catalogue record, OR by linking to an existing one.
As many libraries do this, some records receive more inward links. Not all inward links will go to the same record necessarily – perhaps two or more key ‘nodes’ will appear in the network for a single bibliographic item – but this doesn’t matter, it is ultimately self correcting, and self correctable – afterall, you only need one connection between the two parts of the network for it to be clear that all the connected items are the same.
It also opens up the possibility of explicitly recording that ‘these two things are the same’
I’d stress this doesn’t have to be the WWW – just a linked environment.
You would need to start thinking not just in terms of ‘catalogues’ – what your library has – but also in terms of ‘indexes’ – what do you want to index? To index your library, you would crawl all the records in your library, AND all the ones you link to – you could take this further, and grab extra information from others if you wanted – but the ones that are most linked to, would be the more ‘authoritative’ – allowing records from small, but specialist, suppliers to be more authoritative than those from the traditional ‘authorities’
A Union catalogue could be built by doing a wider crawl – and the ultimate Union by crawling everything. Yes, there would be duplication, but this would be on a smaller scale than currently – and the more links you add, the less duplication there is (more data)
Although I’ve approached this in what is perhaps a Library-centric way, a linked environment allows links to anywhere – so you can link to (and index) resources outside your domain – building an index to serve your users. –standards might be needed – but in some ways links would provide aspects of this, crosswalks could be built from practice rather than theory. Diversity would be embraced and used.
The idea of a linked resource discovery infrastructure requires us to change the way we think about our ‘catalogues’ (whether that be library, archives, journal indexes or other). We’ve treated these as isolated instances – if you have one sheep, you have a pet – you can focus all your attention on it and care for it, but if you put it out to fend for itself, it is going to struggle.
We need to think more like shepherds – treating our resource descriptions as a flock – putting it out to graze, and gathering it back in (via indexing) when we need to.
I’m convinced that Ranganathan would see a linked environment as the next step – the next ‘epoch making contribution’ – he may not have been able to anticipate the information revolution that computers would bring, but he said “What further stages of evolution are in store for this Growing Organism – the library – we can only wait and see… who knows that a day may not come … when the dissemination of knowledge, which is the vital function of libraries, will be realised by libraries even by means other than those of the printed book”. There is no doubt in my mind that he would have embraced the opportunity offered by a linked information environment – and this is what we need to do now.