Magic searching

I have written a review of a new book Magic Search: getting the best results from your catalog and beyond, which will appear in due course in Journal of Documentation. Here’s a flavour of the review of what proved to be of much more interest than we have any right to expect from a book about the sub-divisions of the Library of Congress Subject Headings.

MagicSearch(L)This is not, it has to be said, an immediately gripping book, despite the best attempts of the authors, or perhaps the publisher, to present it with a Harry Potter-like title . Subtitled ‘Subdivide and conquer with LC subdivision!’, it would seem to be a good candidate for one of those competitions for most obscure or nerdish titles of the year. Is there really a book to be written, not even about the Library of Congress Headings – themselves perhaps not the most immediately accessible or popular topic – but just about the sub-headings.

Such cynicism should be stilled, for this short book is interesting in a number of ways.

Most obviously, it will be of value to those – and the books is pretty clearly aimed at library professionals – who search databases indexed with LC subject headings. It gives advice on the best ways if using the sub-headings as a means to cutting down lengthy results lists, and giving just the kind of results needed: as a means of improving precision without unduly sacrificing recall in fact, although this terminology is not used. Of over 3,500 available sub-divisions, the authors pick about 500 which they judge of most value, and show how to sue them to best effect. So, if you want resources with images of people, ‘PORTRAITS or PICTORIAL WORKS’ will do the trick. To find how an artist of writer has been received, you need ‘APPRECIATION or INFLUENCE’. To get at primary sources on your topic, the magic search (sic) involves ‘ARCHIVES or SOURCES or DIARIES or CORRESPONDENCE or NARRATIVES or INTERVIEWS or FACSIMILES’. And so on, for 19 short chapters on different subjects and topics.

The book then will certainly be useful in improving searching practice. But its interest goes beyond this, in that it illustrates some of the current issues in bibliographic retrieval today. For one thing, there is a strong Google influence throughout the book; perhaps surprising considering that the LC subject indexing methods might be considered the epitome of the ‘librarianly’ approach. The authors, however, point out two interesting facts. Google Book Search, which – whether we like it or not – is going to gain a very significant position over the next few years, will be making use of WorldCat metadata, and hence be searchable by LC headings, and their magic subdivisions. And the authors show how the subheadings can be used to form Google-like searches, providing, in a sense, the best of both worlds.

Of course, this happy situation depends on a number of factors, as the authors remind us. It relies on the continued practice of intellectual indexing using controlled vocabularies, and a powerful and well-argued plea is made for this. It relies on the Library of Congress getting their vocabulary right, and some good recommendations are made. And, of course, it relies on searchers knowing what they are doing, and having appropriate interfaces, to catalogues and databases, to support them. When a book of this sort has to be researched and written to remind the professional library / information community of the right way to use these tools, one wonders how much use is being made of them by most users of catalogues and other bibliographic collections. This book is a very good start, but a major task of consciousness raising awaits.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s