Preserving the random with coarse-grained filters in Twitter
One of the frustrations people have with Twitter is its simplicity. Twitter is an authoring environment for hypertext limited to 140 characters and a method of publishing and subscribing to an almost unlimited combination of social graphs. It achieves some complexity through its API, which allows it to be mashed up with other applications. In this sense it adheres to David Weinberger’s idea of “small pieces loosely joined.”
Twitter’s simplicity means the barrier to getting started is very low, register an identity, type 140 characters and click “update.” Understanding the value of Twitter doesn’t come until later. Non-users and new users can’t actually experience Twitter. The public timeline is there as an example, but to generalize and form opinions based on this evidence would lead one solidly in the wrong direction. The public timeline could potentially be decoded, but it’s a task very similar to spending time with Humphrey Chimpden Earwicker, Anna Livia Plurabelle and dream logic of Finnegan’s Wake. Or as James Joyce put it: Here comes everybody.
Veteran users of Twitter experience something very different from the public timeline. And it’s those regular users who begin to long for more controls, more features to help them refine their Twitter experience. Generally this is expressed through a desire to configure and define groups within the larger pools of the followed and the followers. By concisely defining groups a Twitter user could get exactly what she wanted.
But getting “exactly what you want” is exactly what you don’t want. Fine grained controls and filters are generally used to focus on common interests and concerns. The result is pre-defining the message flow you receive, creating an echo chamber. Random and negative feedback have an important role the health and stability of any dynamic organic system. When Twitter only brings you what you expect, it loses its value.
Twitter will grow new features, all applications do. But what if, rather than think in terms of precision, exactness and clarity; we thought of coarseness, randomness and ambiguity. What kind of coarse grained filters would preserve the random in a users Twitter stream? The seed for this rumination was inspired by a conversation on @Newsgang Live about squelch as metaphor for filtering Twitter streams. Imagine filtering the stream based on frequency of tweets, or location of tweets. By tuning into quadrants of the Twitterverse with coarse-grained filters new voices could be discovered. So often we think in terms of signal versus noise, but when we think of noise perhaps we should take a lesson, and listen with the zen ears of John Cage.