The words chosen to describe my job are varied: sometimes information professional is used or perhaps the more traditional variants of ‘librarian:’ subject librarian, academic liaison librarian, research librarian (which is my actual title) or perhaps a hybrid label such as information technologist is used. Descriptive discourse and word choice are critically important. The concepts which they engender should reflect (in this case) the duties and responsibilities of the role and the changing nature of the profession.
As someone who shepherds the research collections in one of the largest open-access humanities libraries in the United Kingdom with some 4 million items, my library is not immune to the dynamic digital world and its impact upon the traditional printed publication output across the humanities. One of the key elements in any library organisation is its classification system (i.e. how the books are arranged) so that they can be stored and retrieved. There are various systems (like the Dewey Decimal, Library of Congress, Bliss, etc.) which can be used but the thread of commonality across the systems is that they collocate similar materials based on hierarchical subject trees. There are standards which libraries use to catalogue material so that if you use a public or academic library, you will be able to search across the ‘author’ or ‘title’. If you were researching a known topic you could even use the ‘subject’ search to restrict to the specific language used in cataloguing materials. For example, criticism of Shakespeare’s work in a particular historical period can be found using ‘Shakespeare, William 1564-1616 — Criticism and Interpretation– History’. Keyword searching using terms like criticism, history and Shakespeare might bring back some of the results but a subject search using ‘controlled vocabulary’ like the Library of Congress Subject Headings (LCSH) brings a more precise result (Figure 1 below).
I don’t have a problem with keyword searching as long as there is a manner to use ‘controlled vocabulary’ for precision-searching. Keywords just don’t cut it. (It’s the same problem I have with Google and the Google Scholar.) Unfortunately a ‘keyword’ default in most catalogues is the norm. From the results list one could then pivot using the LCSHs (just like the above search). However, many of the ‘next-gen’ library catalogues now remove the possibility of using a ‘subject’ search using controlled vocabulary. One can use a ‘subject’ filter/facet (which is a form of keyword filtering) (see Figure 2).
Comparisons of searches between the two methods should achieve identical results if the algorithms are working correctly. Clearly the 95 and 240 are divergent and evaluation of the former’s results is more accurate and precise even when aggregation of all of the time periods are taken into account. So the search methods reveal that a proper ‘subject’ search, in fact, produce different results and ranking from the keyword ‘subject’ facet. This frustrating conundrum demonstrates the importance of subject hierarchies and the problems with keyword (uncontrolled language).
What is the solution?
Researchers need to be aware of some of the mechanics behind search and the role that controlled vocabulary plays, particularly hierarchically. Think of it as a pre-emptive strike in the battle to organise information (for yourself). Word choice may inform initial keyword choices but researchers must navigate towards a hierarchical structure of order (whether it is the index or taxonomy or subject/hierarchical term structure) within a database. Not only is it more precise it will save you time.