D-Lib Magazine, November 2006, Volume 12 Number 11
Beneath the Metadata: Some Philosophical Problems with Folksonomy – Elaine Peterson
Elaine Peterson has written a very interesting article looking at some of the philosophical contingencies of user tagging of folksonomy, something which I think is very necessary and I was happy to see.
In her introduction she makes the important point that, regardless of what one’s metaphysical or ontological beliefs are, for classification purposes, Aristotle’s law of non-contradiction must apply. A book cannot be both A and not-A if we considering its classification. She rightly notes: “It is irrelevant that digital items can reside in more than one place, since one is talking about a classification scheme, not about the items themselves.”
However, I disagree with her second introductory point, that the cataloguer, like the librarian, should be neutral and follow the intent of the author. It is questionable, in the first place, whether we can even access the intent of the author (see Foucault, Barthes, Derrida). Nevertheless, as cataloguers and librarians our charge is not the book, but the user, and we should look at things from their perspective. Our “neutrality” does not come from following the author–creation mythologies are not in science, they are in religion; Hannibal Lecter’s guide to kidneys and fava beans is not under culinary arts, but deviant psychology. Thus Peterson’s assertion: “the goal is to recognize the author’s intent over others’ interpretation,” is flawed. There is a social aspect to classification: no matter how much the author wishes or believes, a fictional account cannot be placed in the non-fiction section.
She then goes on about folksonomies, noting their oft repeated benefits and detriments, and makes the claim that “philosophical relativism appears to be the underlying philosophy behind folksonomies.” Beyond synonym and typographical errors, it is possible that users might tag an item with directly contrary tags, violating Aristotle’s law of non-contradiction. Her example, a picture of a white horse could be tagged both “black horse” and “white horse.” In her view this is detrimental to searching and retrieval.
The critique I can give to her position is that classification schemes give only the illusion of order and hierarchy where it doesn’t truly exist. We all know that classifications have their own cultural–i.e. relativist–biases. What we don’t like to admit is that these classification systems come from not only group discussion and agreement amongst professionals, but also the cultural environment where the system is developed. And societies will cover over the fact that the classification systems used to define its reality have no fundamental basis (Foucault). Peterson doesn’t consider the possibility that reality is relativist, thereby making “traditional” classification are static, possibly not reflective of current cultural beliefs, or even the tool of dominate power structures. This is not something I am arguing is true, mind you, but traditional classifications do have their own philosophical problems.
Which is exactly why controlled and natural vocabulary classifications should live together, within one bibliographic record. This is all possible within the OPAC. Not only for the fact that folksonomies can highlight the inherent biases within a classification system, but for the very fact that cataloguers make mistakes, they miss classify, or they miss a subject heading which could apply. And, at the same time, controlled vocabularies can serve as a stable reference point.
However, I think Peterson’s article raises some excellent points for discussion, and I recommend you read it.