Saturday, February 02, 2008

On Tags and Signs: A Semiotic Analysis of Folksonomies

In the last few years, the practice of tagging resources on the World Wide Web has become a more popular activity, due in part to the success of websites that feature tagging functionality prominently, such as, Flickr, Technorati, CiteULike, and LibraryThing. These websites allow the collection and sharing of bookmarks, photographs, blog postings, academic papers, and books respectively, and tagging is a key method to enable retrieval and sharing within these sites. Delicious offers the best description of tags and tagging, and this description can be extended beyond bookmarks to a general definition:
Tags are one-word [or multiple words written without spaces or with underscores] descriptors that you can assign to your bookmarks on to help you organize and remember them. Tags are a little bit like keywords, but they're chosen by you, and they do not form a hierarchy. You can assign as many tags to a bookmark as you like…This is great for organizing and finding personal data, but it goes even further when someone else posts related content using the same tags. You begin building a collaborative repository of related information, driven by personal interests and creative organization. (
Tagging information as a practice existed prior to these websites popularizing it, but these websites were among the first to extend the practice beyond the domain of content creators or information specialists to the general public (Weinberger, 2007, p. 92). When tagging is done by the public, it is often known as a folksonomy (
For an understanding of how folksonomies work, or often do not work, the academic tradition of semiotics offers help to “understand what goes into a message…[it] also help[s] to understand how the message comes to have meaning” (Littlejohn & Foss, 2008, p. 105) Through a semiotic analysis of the characteristics of folksonomies, I will explore three main difficulties inherent to them. Despite these difficulties, they are emerging as a popular new form of communication in line with the more fluid notions of meaning consistent with the post-structuralist notion of semiotics. The characteristics to be examined are that, first, folksonomies’ collective nature facilitates open meaning, second, that tagging for the self results in confusing connotations, and third, that there is a lack of message coding (in the semiotic sense). Prior to discussing these characteristics, however, I will provide background information about the problems inherent to the use of folksonomies, and insight on how semiotics provides a useful conceptual tool to examine them.
Problem with Folksonomies
The practice of tagging is useful as a personal means to organize data and as a mnemonic device for one’s own retrieval. Indeed, it was originally for this use that Joshua Schachter created (Weinberger, 2007, p. 162). When one moves beyond the personal uses for tagging, however, to the social aspects, the utility of tagging becomes more problematic. There are essentially two levels of complications: 1) people’s tags may be difficult for others to understand, 2) people may have tagged items inappropriately for others’ needs. A good example of this problem was a recent search of Flickr for photos of Hawaii. In addition to the expected images of beaches and volcanoes, one also gets, among the top results, a picture of a couple at an outdoor restaurant (presumably located in Hawaii) and a close-up of a lizard (also presumably located in Hawaii). Another example of tagging difficulties arises in a scan of the most popular tags at, revealing some more meaningful tags (youtube, tv, psychology), but others that are puzzling (toread, fic, ubuntu, fun). People are participating in folksonomies, but not always doing it in a manner that is useful to others.
Use of semiotics
Semiotics is the “study of signs” (Littlejohn & Foss, 2008, p. 35) and tags are definitely signs. Using a triadic structure of a sign with a signifier, that is the “material dimension of the sign,” a signified, “the cultural or conceptual dimension,” and the referent “the real thing in the world,” the structure of a folksonomy sign can be broken down (Black, 2007a, p. 5):
  1. signifier = a word or single term that is a link appearing on a webpage
  2. signified = what that word link represents to the user, i.e. an expectation of resources on that topic
  3. referent = the actual resources that have been tagged with that word, at that point in time
Examining folksonomies through this semiotic structure initially reveals one of the fundamental difficulties with folksonomies, that is, the referent is almost always unknown to the user until he or she clicks through. Often with signs, there is a clear or discernible referent, yet folksonomies, by their very nature, are a complete mystery to the user, making a single, clear meaning unlikely.
Another problematic element is that current technology and practice do not disambiguate the multiple senses one signifier may have (for example, synonyms such as: asp, the programming language and asp, the snake of Cleopatra fame) or to join multiple signifiers sharing the same signified (for example, blog, blogs, weblog). Morville believes that folksonomies are thus fundamentally flawed due to their inability to “handle equivalence, hierarchy, and other semantic relationships caus[ing] them to fail miserably at any significant scale” (as cited inWeinberger, 2007, p. 166). However, Weinberger believes that computers, possibly via artificial intelligence, will eventually address some elements of ambiguity in folksonomies (Weinberger, 2007, p. 166).
For Saussure, a founder of semiotic theory, “the meanings of a sign were fixed socially by convention and, thus, were independent of any consequential variation in interpretation” (Danesi, 1999, p. 11). Yet, consequential variation in interpretation is often widespread with folksonomies, indicating that the meaning offered by tag signs is more complicated, compared to other uses of signs.

No comments: