By Dave Hook I don’t shy away from opportunities to geek out. Yes, I’ll happily…
Good examples of folksonomies are found in the hashtags used on Twitter, which were originally an innovation of Twitter users. Hashtags are abbreviations, words or phrases preceded by a crosshatch symbol, used to index messages or “tweets” on an author’s microblog. Messages sharing a hashtag are aggregated. If large numbers of messages share a hashtag, an algorithm ensures these messages have higher visibility or “trend.”
Tags or index terms can only be added by the author, unlike Flickr which allows users to tag each other’s photos. Twitter tagging is a flat classification, that is, there is no relationship between index terms. Twitter tags are not chosen from a controlled lexicon, but over time, some standardization has taken place so that, for example, #cdnpoli is consistently used for messages related to Canadian politics.
Twitter hashtags may comprise more than one word, but there cannot be a space between words, for example, the hashtag #OccupyWallStreet aggregated tweets about social protests in 2011. Hashtags are simple to apply, reflect the author’s viewpoint, and when used frequently, can ensure that messages become” trending topics” which are highly visible on the site and may be included in Twitter’s annual review.
Twitter folksonomies have expanded the uses of classification in ways never imagined by librarians. One of the unique uses of hashtags has been to create games. For example, using the hashtag #LessAmbitiousBooks, users gave books new titles, such as “Where The Mild Things Are”. Hashtags are also used to convey sentiment such as sarcasm, humor, or pain, for example, “Trying to be cool #fail” or “My cut-up legs from wearing shorts playing softball look so good #sarcasm.”
Hashtags share disadvantages found with any folksonomies.
Using synonyms will prevent the aggregation of all tweets on a subject, for example #OWS was also used during the 2011 protests, and was separately filtered. Misspelling or alternate spellings of a hashtag also prevent aggregation, for example, #HonorTheApology instead of #HonourTheApology. Ambiguous words limit the utility of hashtags, for example, #cloud aggregates messages about meteorological phenomena and remote server based computing. Finally, there is no way to classify broader, narrower, and related terms, such as #Food and #SchoolLunch, or #ClimateChange and #GlobalWarming.
Twitter allows for full text search, but hashtags do help users find specific subjects. For example, searching for the word “apology” in messages will not aggregate only messages related to the federal government apology to indigenous peoples. These messages are tagged with “HonourTheApology” and if searched using that hashtag, other messages containing the word “apology” will be filtered out.
Twitter hashtags are evolving into controlled vocabulary but will remain a flat classification system with other inventive uses.
Mary Kosta is the Archivist for the Congregation of the Sisters of St. Joseph in Canada. She can be reached at mkosta [at] alumni.uwo.ca.