Today’s comic from xkcd is titled “Ablogalypse.” Randall Munroe shares his work under a Creative Commons License, so I’m reposting it below*:
Three things about this chart:
1. Notice that mentions of “blog” haven’t declined much. People are still blogging. People are still talking about blogging. But people are also finding new uses for tumblr sites, and many of those uses are absurdly shareable.
2. We saw a similar process a few years ago with social network sites. Several public commenters looked at the rise of facebook and suggested it meant a decline of blogging. Chris Bowers (I think, can’t find the post) responded that instead we were seeing finer-grained niches. Blogs used to be the only self-publishing game in town (’01-’04ish). So early adopters used blogs for all sorts of communicative purposes… even ones which a medium designed for instantaneous default-public, default-permanent writing is poorly-suited. As the social web has developed, new platforms have been created with different affordances. The more sophisticated users have started to select the right tool for their communications purpose.
3. Tumblr sites are particularly good for fun viral stuff. Last week’s phenomenon, Texts From Hillary Clinton, is a great example. Two netroots politic0-types came up with the lolcats-style idea over beers. A few years ago, they would’ve launched it as a blog. That would’ve worked alright, but blogs are a little clunky if you just want to post images and short commentary. So today they use a tumblr site instead. To the extent that images-and-captions are more viral-friendly (or “upworthy“) than their text-heavy equivalent, we ought to expect a spike in tumblr’s google rankings.
Last Friday at the Midwest Political Science Association Annual Meeting, I gave a presentation of my latest paper, Social Science Research Methods in Internet Time. It’s essentially an extended rumination on the phenomenon associated with this graph. New features of the social web emerge fast. It creates a novel research problem — our most robust social science methods are based in the ceteris paribus assumption that the communications network we sample at time X will be basically the same as the network in existence at time Y**. I argue that, in the face of the ongoing adoption and adaptation practices, our best research options often involve embracing the messiness, being transparent about our data limitations, and hacking together kludgy research designs that provide some analytic leverage on how the system is evolving, and how it all fits together. …In light of this week’s comic, maybe I should have added “keep a sense of humor” to that list.
*Please tell me you’re already regularly visiting XKCD. New comics come out on Monday, Wednesday, and Friday. There is no such thing as a person who would enjoy shoutingloudly.com but dislike xkcd.
*Where X = the time when you conduct the research and Y = the time when your research is published. It’s a reasonable assumption most of the time, and hellishly problematic when it proves unreasonable.