Wednesday, 13 June 2012

Random thoughts on subject headings in library catalogues (warning: full of cataloguing jargon)

At Cardiff University, we allocate Library of Congress subject headings (LCSH) to our catalogue records. This was not always the case: before we acquired our current web-based LMS in 1999 we did not do this, but with our LMS came lots of records from other libraries which did, and we also took part in a collaborative cataloguing project at around the same time which asked for them. Since then we have gone with the flow and tried to use them, with varying degrees of consistency.

I had used LCSH before, but that was many moons ago, when basic records of books were added to a computerised system for circulation purposes only and the catalogue was a separate beast entirely. The structure of LCSH is designed as far as I can see mainly for either a card catalogue or a printed reference book. The headings do not really make much sense when not seen in hierarchical form, and it's unlikely that many users actually search them in the way originally intended (browsing them in some sort of linear progression).

In the period between the card catalogue and the web-based OPAC, we did have a basic keyword search function, and over time the practice was to add simple (but completely uncontrolled) keywords to the records. As the system would not search words which occurred within the title of a book, very often these words were the keywords (now redundant, though many are still there). I no longer add them except in the rare cases where I really feel that only a British English word (or even a Welsh word) will be useful to our own users and the word in question does not appear in the title of the book. Sometimes I wonder why we go to so much trouble to find the right LCSH which, when found, turns out to be a term none of us would ever use. On the other hand, there are dangers in the freedom of the uncontrolled keyword too. Our older collections contain works couched in language which would not be used today, and in the past these books were sometimes catalogued by people who were not aware of various nuances and shibboleths and general opportunities for causing offence through choice of words. (I think I have now found and removed all the "610.20 Church of Rome" headings which I came across, likewise "Papists", and various other keywords suggested by the titles of the works of forthright early Protestant clergy).

As we move to the next stage, the new generation "resource discovery platforms" which are replacing the more structured catalogue, I wonder whether LCSH still fulfils its purpose, or whether we should once again be looking at something less rigid. Ideas, please!


  1. We use quite detailed subject headings in our catalogue, though not LCSH. Partly because we're a single subject research library and there's a need to drill down further into the content in order to distinguish one economics book from another! It's all about increasing discoverability (or what in the old days would probably be called 'access points'). We also use the headings as a basis for subject updates and guides. But whether there really is still a need for controlled terms as opposed to less rigid keywords, I'm not sure...

  2. Like you, we did not add LCSH to bib records, and until quite recently just accepted the headings which started to appear in records imported via our LMS. About a year ago, after looking into the new discovery systems, we decided that in order to improve "discoverability" LCSH were a must, so we now routinely check, amend and auhtorise them if they appear, or add them if they don't!
    We've recently begun a retrosepctive programme of adding LCSH to those bib records that don't have them. This is a mammoth task as at the last count there were 136362 such records!
    So, I think I'd conclude that I'm rather more in favour of LCSH (or other controlled subject headings) than I am of user-assigned tags or uncontrolled keywords.
    I suppose the only proof we have of usefulness is whether or not users find them helpful in locating relevant resources.

  3. In an admittedly long, long ago time. I was trained in LCSH during my degree. Thinking back I believed that LCSH formed in addition to the subject heading/keyword discoverability aspect, an alternative shelf location tool to good old Dewey. As opposed to a purely controlled language subject list.

    If this is the case then the hierarchical nature of the headings would be vital.

    As I say since then I have always used faceted schemes or the standard UDC and DDC, so I am not sure if I am misremembering LCSH's purpose. (and occasionally I can be wrong but don't let anyone know);-)