Twittering Machines

Twitter Visualized – Scott Templar

Any excuse to post a picture, and reference, one of Dataisnature’s favourite Paul Klee drawings is always welcome. Dataisnature/Mr Prudence/Transphormetic is finally twittering on a tiny branch of global Twitter tree. You’ll find links to sites outside of the standard D==N subject matter, as well as found aphorisms and other pithy remarks, updates and the de facto hour to hour miscellany. Social networking sites increasingly raise the question of identity(s) Which of the many different me’s are allowed a say in a specific networking channel?. Will one identity take over and become a monad over the nomads in the datastream, or will the personal, impersonal professional, irreverent, inane and metaphysical get all mushed up into an info-noise soup? Who knows and does it matter? For now, at least, rather than having a personal as well as specific Dataisnature Twitter account, ill just have the one. Or else I might end up more overloaded than before and finish up looking like the exhausted, exasperated bird on Klee’s machine.

Not surprisingly Twitter is ideal feed for the data visualisers. A quick bit of googling reveals a mass of visualisations ranging from the poetic to the mandatory – the best to my eyes, are the ones that reveal the temporal side of interconnectedness such as Socialcollider.

3 Responses to “Twittering Machines”

  1. toxi writes:

    Hey Paul, I just looked at Scott’s “twitter visualized” & commented on his blog, but since you linked to it I feel this needs some explanation here too: With any dataviz project I find it quite dangerous if people make claims to visualize A, but for various reasons end up visualizing B, maybe even without realizing. In Scott’s case his claim of visualizing a twitter network is simply plain wrong, since what’s visualized is nothing more than the HTML markup structure of a twitter profile page. For anyone with more than 36 followers this will be absolutely the same (because of the fixed twitter page template), so a Stephen Fry graph will be only differing to my own in the node cluster related to his actual message feed. The cluster relating to contacts is the small one with blue & purple nodes and will be limited to 36 nodes and be the same for Mr. Fry and myself, even though he’s got 485 times more followers than me (395000 vs 810)! I call this whole thing “dangerous” since this example just shows again how easy we’re to pick up memes (dataviz) and also tend to believe images more without questioning. If dataviz is becoming more of the norm and new lingua franca to present all sorts of data, then there’s a huge responsibility involved on the designer’s/author’s behalf…

    My 2 pence.

    Ps. Also thanks for the socialcollider plug! This one has issues of its own, but IMHO less conceptual ones & more related to actual lack of features to help the current presentation…

  2. Maxwell's Demon writes:


    Why English social fears did not stop me joining twitter and starting to like it.

  3. paul writes:

    Heh Toxi,

    Thanks for dropping by and elucidating some of the pitfalls and unintended misconceptions when attempting a visualization or claiming to have visualized a data set. For my part I hadn’t delved too deeply below the bonnet of Scott’s visualization to see the issues you mentioned – so thanks for the great detective work. Data Vis takes up a very tiny amount of space on my blog and I wouldn’t claim to be an expert nor overly interested in this specific area compared to other areas of digital creativity. Having said that your SocialCollider peice has that rare blend of great aesthetics and story telling without being obtuse or difficult to understand.

Leave a Reply