Problematic Solutions to Controlled Vocabularies……

Controlled vocabularies in modern classification systems were designed to assist in making information more accessible to library patrons and professional researchers. The very idea that vocabulary can be controlled finds its origins in Enlightenment ideology in which modern science was founded on the principles of ‘knowledge and truth’. Within the field of linguistics, this meant that the greater an understanding of the mechanics of language and the greater the accuracy and control over words, the closer one could come to truth. In applying this ideology to library classification systems, both the Dewey Decimal System and the Library of Congress Classification System attempted to create subject headings under which library content could be listed. The claim that these systems are biased is no longer really in dispute as both systems have attempted in recent years to adapt to new emergent disciplines that have been marginalised due to the biases within the respective classifications. However, the continued attempt to regulate the classification of information through slightly more ‘flexible’ controlled vocabularies is detrimental to real innovation and creativity, not only in academic and scientific research, but also in terms of promoting real diversity and creativity privately and publicly in social, political, economic and cultural spheres. It is ironic that the very systems that set out to improve accessibility to information and to thereby foster greater innovation and awareness, has in fact lead to greater ignorance. The problem persists in so far as classification systems continue to be regulated by outmoded ideology in which classification ‘specialists’ take as their starting point the mantra that information can be classified in a coherent and organised way under specified subject headings. The truth is that until classification systems more fairly account for the differences and diversity that exists within singular texts, then these systems will continue to be biased, acting as an obstacle to knowledge as opposed to a medium to accessibility. The following paper sets out to explicate the problems created by controlled vocabularies in classification systems by discussing the issue from a poststructuralist perspective, specifically utilising Derrida’s work in Of Grammatology to explain the contradictions in traditional classification systems and to further critique some modern solutions to these problems.


Borges (1952, 104) asserts that “there is no classification of the universe that is not arbitrary and conjectural”. Borges identified that at the core of any attempt to classify or to organise objects through controlled vocabularies is the fact that this organisation is based on shifting premises that can be challenged from a point of difference. Beghtol (1986 ) illuminates the problem with traditional classification systems further by arguing that these systems are created through concepts of authority, status and control. In this sense, traditional classification systems become homogenous and hegemonic, leading to theorists like Shirkey (2005) to argue that classifications are, by their very nature, biased. Miksa (1998, 81) demonstrates that the problem with classification lies in its Enlightenment roots, arguing that classification is based on “the idea that somewhere, somehow, we can, or should try to, produce the one best classification system that will serve all purposes”. What is emphasised here is that there is one purely scientific system that pertains to truth in that this one system is complete in its ability to categories all knowledge. Miksa (1998, 81) goes on to highlight the assumption that “knowledge categories are by nature hierarchical and logical in a classical, systematic sense”. Any kind of hierarchy is established to deny real difference in a subject because everything that exists under that hierarchy must be shaped to fit into categories that the hierarchy dictates. If that hierarchy is Western, or North American, then there is naturally going to be a bias towards the ideological prioritisation for those demographics. Such systems, then, become too rigid and cannot adequately account for emerging disciplines or in fact, for the transferral of information across national and continental boundaries.

Theorists have developed the critique of traditional classification systems further to incorporate contextual elements into the debate. These contextual arguments all revolve around the idea that classification systems work to identify similarities between objects and to thus categorise them under related headings. Bowker and Star (1999, 131) suggests that “classifications that appear natural, eloquent, and homogenous within a given human context appear forced and heterogeneous outside that context”. Thus, extracting classifications from their original context within the system demonstrates just how biased the system actually becomes. Mai develops this concept further to highlight the prominence of ‘similarity’ in classification systems: “most bibliographic classification theory stipulates that documents are holders of concepts and concepts are context and human independent constructs and that classification brings together concepts based on similarity”. Lakoff (1987, 6) explains this idea of similarity further by explicating that since the writings of Aristotle, and following through the entire history of Western thought, objects were categorised based on whether or not they had ‘common properties’. Olson (2001, 116) develops the concept of similarity more to incorporate concepts of ‘sameness and difference’ in organising information: ‘once we collect this innovative material we try to organize it by gathering what is the same […] We build our classifications using these facets that bring things together according to some kinds of sameness’. However, Olson, like many contemporary theorists navigate more towards ‘sameness’ as a regulating principal of classification systems, ultimately, paying tribute to the dualism of difference but never seriously considering its significance comprehensively enough. Olson further relates the idea of sameness of that to ‘discipline’, referring to discipline as ‘the primary facet in our classification systems’. ‘Discipline’ is a word that implies an authority and strictness over controlling vocabularies and categories of information.

What is interesting is that any attempt by theorists to explain classification as biased through the concept of sameness, implicitly means using the concept of ‘difference’ as a critical hinge upon which to base those critiques. It is through this ‘hinge’ that solutions to the biases within controlled vocabularies start to emerge. Clare Beghtol (1982, 2), for example, suggests that “increasing multidisciplinary knowledge creation makes it critical to reconsider the traditional reliance on discipline-based classification and to try to solve the problems that orientation has created”. Olson (2001, 120) continues to highlight, in relation to English Literature classification, that this traditional mode privileges ‘colonisers over colonised’. She (2001, 118) further develops the importance of ‘difference’ in that she explains that “Recent recognition of the validity of oral literary traditions and the questioning of existing literary canons suggest that this definition of literature is exclusive rather than inclusive. It is defined by difference as much as by sameness”. It is through the importance of understanding difference in classification systems that some solutions to the biases begin to emerge. Olson’s solution appears to be more practical and achievable than many others. The problem is that Olson’s solutions (2001, 120-122) self-consciously lack real difference in suggesting change, preferring to base her solutions around ‘local control’ rather than any real radical changes. This means giving libraries, both regional and national, ‘notational options’ that allows them to make amendments to the subject headings so as to find vocabularies that are more suitable to the given context. Olson also suggests that “Flexibility can also be achieved by varying the citation order of classifications – shifting which samenesses get priority. It must involve rejecting at least some of the samenesses and differences of our classifications.” Olson justifies these changes by referencing postmodernism’s rejection of universals, seeing more local control as a disruption of traditional all-encompassing systems. However, she fails completely to really understand postmodern and thus poststructuralist concepts of difference and disruption. The suggested solutions simply replace one form of universal control with another more local version of the same thing. The solutions still rely on ‘specialists’ to assert authority over the vocabularies used to classify, taking the authority out of the hands of the users and placing it into the hands of individuals. It is as equally problematic as Mai’s (2010) concept of ‘cognitive control’ which allows for the continuation of the traditional system once it has been properly theorised, questioned and explained, so long as it is self-conscious to its own potential biases. The problem with this is that being self-conscious of bias does nothing to eradicate the bias, it is simply bias in a softer guise. And in any case, no matter how self-conscious we are of the bias, objects in our libraries will continue to be irretrievably buried under inadequate subject headings.


What becomes clear, then, is that there is an awareness of poststructuralism’s influence through the referencing of the concept of ‘difference’ in attempting to understand and solve the problem of controlled vocabularies in classification systems. However, there is also a clear reluctance to engage with poststructuralism in a meaningful way. In fact, there is a clear misunderstanding of poststructuralism in the dualism of understanding ‘difference’ as the opposite of ‘sameness’. This suggests that there is an assertion of poststructuralist politics in promoting some ‘difference’ at local and national level, but that this politics is built on structuralist rather than poststructuralist linguistic foundations, thereby rendering it contradictory and self-defeating. All of the ‘solutions’ to classification bias are retained within a traditional mindset in which texts ought to be categorised into similar or related categories that are hierarchal in nature. No matter how much flexibility one allows within this model, or no matter how much ‘trustworthiness’ is achieved due to cognitive control, there still persists a traditional model that gives authority control to an oppressive few. The remainder of this paper will attempt to explain the real value of poststructuralism to this debate and will further attempt to demonstrate the radical potential of poststructuralism to not only disrupt traditional classification systems, but to disrupt them in a positive way that could lead to more meaningful solutions to the problem.

Part two, explicating how poststructuralist theory impacts upon classification systems, is coming soon…………

Leave a Reply

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

You are commenting using your 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