User:Dcoetzee/Named topic bias
The named topic bias is a particular natural bias possessed by all encyclopedias: a tendency to create more articles — and more detailed articles — on topics with a single, widely-agreed upon name. Some obvious examples include:
- People, who usually have a single "real" name which they are widely known by (George Washington)
- Nations, states, provinces, counties, and cities/towns such as France, British Columbia, North Dakota, Beijing
- Major fictional works such as novels (Gulliver's Travels), songs ("American Pie"), television shows (Buffy the Vampire Slayer), and films (Jaws), which are given specific titles for the purpose of marketing
- Programming languages, wars, companies, products, websites, animals, and so on.
There are exceptions to these which cause contention, such as people with two widely-known names (Lewis Carroll/Charles Lutwidge Dodgson), battles or skirmishes with no widely accepted historical name, and so on. On the other hand, we have relatively few articles on topics which do not have any widely used name, even if it is a specific and coherent topic worthy of discussion. Some existing examples of this include:
- Proof of impossibility
- Superhero live-action television series
- Many current event articles such as 2006 definition of planet
- Some of the more creative lists and categories.
A large proportion of new articles that are deleted are about well-named concepts, such as vanity or attack articles about bands, people, companies, websites, or works of fiction. Almost never do people create new articles about events such as particular games in sports or a particular speech given by a political figure; although these are widely-known topics, they simply don't have a name. Articles based on original research almost always have a title which is also original research, a neologism or invented phrase.
In mathematics, some of the most inspired developments resulted from the identification of an underlying abstract structure shared by many disparate concepts in different subfields. The concepts studied by abstract algebra and topology are particularly pervasive, as well as matroids. I believe Wikipedia should draw on this example and create articles which unify diverse ideas or topics in compelling ways.
As an encyclopedia, we are not in a position to coin new names for concepts; this would be original research. However, as the above examples show, it is not necessary to coin a name for a concept as long as we can provide a brief, plain language description to use as an article title. Articles whose titles are neologisms or not widely used are often recommended for deletion at Wikipedia:Articles for deletion, although the concept behind the name may have integrity; often a move to a "more boring" name is more appropriate.
The purpose of this essay is threefold:
- To provide a name for this bias and so make it more available for discussion (an exploitation of the named topic bias itself);
- To provide a source to cite when addressing discussions of topics that have conceptual integrity but no clear name, such as deletion or move discussions;
- To drive expansion of Wikipedia in the area of topics with no singular, widely-recognized name.
Note that this problem isn't necessarily endemic to wikis; there are wikis such as TVTropes that invert this tendency by strongly encouraging users to identify patterns and coin new names for them.