Murder Mystery on Twitter

We’ve seen Twitter haiku and short fiction, but last night user @AustinLugar on Twitter set up a murder mystery told entirely through tweets. You can read it here. It went live last night at 7PM EST, and I managed to catch the first part although I couldn’t stay around to see the conclusion.

I’ve seen a lot of buzz about this type of thing as a kind of “new fiction”, but it isn’t really. The murder mystery genre certainly isn’t anything new, and although it’s being adapted to a new medium, it bears many of the same qualities as, say, a radio program with many different voice actors.

When I look at things like this, that’s essentially what I see — an adaptation of an old form into something new. The ending was even a nice tip of the hat to the nature of the medium when the detective character said he was going to reveal the ending right there on Twitter because all of the suspects were following him anyway.

It’s gotten me to thinking about what genres can and cannot be adapted to a medium like Twitter. Some professional sports teams post what are essentially play-by-plays on their Twitter feeds, and there are even manuals out now for how to create Twitter novels. What about music? Could you adapt a musical composition to Twitter? Eric Whitacre created a virtual choir on YouTube, so what’s to stop someone from making a symphony in 140 characters or fewer?

RT and “oldschool RT”

The various types of Twitter Retweeting have gotten me thinking about the dissemination of information on the internet as a form of topic-comment structure.

Back in the olden days of Twitter, there was an organic social construction made when people would copy wholesale someone else’s tweet, post it to their own feed, and attribute it to the original author using the acronym RT, meaning “re-tweet”. Twitter’s developers actually incorporated this into the site’s setup, and now users have the option to directly retweet without having to copy and paste.

However, the “oldschool RT” still exists, and the main way that I’ve seen it used is when users want to add a comment to the original tweet. In the new RT, the comment is implied — that the retweeter agrees (or, if we can cross sites, “Likes”) the tweet enough to boost it with their affirmation. The new RT is the Twitter form of the “Like” option on Facebook and the up-arrow boosting on sites like Reddit, or the elegant use of ^ as a standalone agreement marker in online chat communities.

This signal boosting and adding group affirmation to an utterance is a novel feature of online language use that doesn’t really have a correlate in spoken language. Sure, we can quote people (“Winston Churchill once said…”) but you already have to be relatively famous for this quoting to happen, and frequently the quotes are misquotes or falsely attributed. With features like Twitter’s new RT, we can see the actual utterance being propagated and given collective value.

It’s cool, the Internets.