Here is a short test: how many tags can you apply to an electric kettle? Three? Four? What about fifty-two? This is how many I counted on Amazon for one standard kettle. A review of the tags demonstrates the chaotic nature of the process. Amongst descriptive tags many are equivalent terms (kettle/ kettles, fast/quick). Others reflect the idiosyncrasies of personal judgement such as ‘good boiling sound’ (a dimension of ‘kettle-ness’ I had not appreciated).
The reason for this pre-occupation with tags is our current challenge of constructing the front-end for the Collections project. We are working on a WordPress website that combines a blog about issues related to OER research methods, and links to resources. Our aim is to highlight licensing issues and encourage the academic community to review methods resources to strengthen trust in the quality of the materials.
Tagging: Between Information Architecture and Personal Information Management
The two aspects of site – the blog and links to resources – raise the issue of how best to make use of tags. Tagging sits a little uncomfortably between information architecture and personal information management. The former is concerned with structuring information on the web to make content findable and favours taxonomies and controlled vocabularies whilst the latter focuses on the individual organisation of information and is idiosyncratic and personal in nature. Tagging represents the point at which these two interests come together. The tags on some sites may only make sense to the person who has applied it. In the Amazon example above there are tags such as ‘Dave’ which may act as an aide-mémoire but communicate little to any other user. Blogs rely on contributors applying tags that both make sense to other users; more than this, the person applying the tag needs to think about the sorts of terms user might search for their piece under.
Taxonomies, Folksonomies and Tags
The blog for the Collections project focuses on research methods and OERs within the social sciences. As an academic site the language used is a specialist one so the onus is on contributors to provide tags that are meaningful to other academics. In some ways this should be easier than with a general blog because content and meaning are more specific. However the field of research methods (even within the social sciences) is full of disputes over language and different disciplines may use different terms for the same method. Since one of the aims of the website is to make OER materials more findable, should we control the tags used to describe the materials or does that go against the spirit of a blog?
Tags have provided the technical means of allowing folksonomies to challenge the authority of traditional taxonomies and some argue that the hierarchical approach of taxonomies are now out-dated. On the other hand part of becoming a member of an academic community is being enculturated into the language of that discipline. Arguments about the meaning of concepts and their relationship to other concepts is an essential part of the social sciences but shouldn’t be seen as a reason to have no structure at all.
Possibilities for the Collections Website
One strategy we are considering is to include the social science taxonomy developed by the National Council for Research Methods in the section of the website that provides guidance to contributors. A simplified version has been used to categorise other aspects of the website such as the video resource. An alternative is to take advantage of the various plugins WordPress offers to make the taxonomy more central to the site and be more directive about blog tags.
It would be interesting to hear the perspective of other projects that have had to deal with this issue. What started as an administrative question over how to structure a website has quickly raised issues over academic classification and subject boundaries as well as arguments over user-autonomy. We would like a system that accommodates different perspectives whilst the tags retain their usefulness in locating resources. Whilst acknowledging the imagination of 52 tags for a kettle we would like our tagging system to represent more of a community consensus.