The Digital Archive

It is much easier to store vast quantities of information now, through digital archives, though care needs to be taken, I think, to ensure some use can be made of these stores. A personal favourite illustration of what to generally avoid (though in this case a brilliantly entertaining entry) is “List of lists of lists” on Wikipedia, which clearly exemplifies the site’s constant threat (and the threat of all digital archives) of becoming merely a hypertext vortex of information:


Still, even such unfathomable quantities of information can make for compelling education and entertainment, given the right treatment. I remember being utterly fascinated by World Population Clocks when we were shown them in secondary school, watching the information from all over the world being processed in real time.

Chatbots allow us to simultaneously add information to an archive in the form of dialogue, and to see this mass of information come to life in front of them – a fully engaging experiment!


The digital archive, with its lack of self-consciousness and usually austere presentation, projects an authoritativeness which is difficult to shake – we trust that the information stored within its recesses is pure truth, and forget that humanity with all its error and  is the only way for it to exist. Hence, such faux-pas as the viral (hoax) tweet that the Eiffel Tower lights were turned off for the first time ever after the Paris attacks,


the “8 spiders a year” myth, and Kony2012, among others. Humanity remains at the very core of the digital archive, in whatever form it takes, and when we forget this we fall into obvious and critical traps

the internet as it currently stands (via reddit)


Twitter archival accounts seem the embodiment of all three of these points, possessing the real-time component, the ability to interact and engage with the information stored, and the often patchy accuracy.

Some of my favourites:

@theirishfor breathes life into old Irish dictionaries, reviving people’s love of language and having a blast with linguistic games

@cursedimages isolates bizarre images found from various other digital archives and tweets them without context (a separate account @uncursedimages seeks to ‘uncurse’ these isolated horrors by providing links to their origins)

@everyword, a twitterbot which from 2007-2014 tweeted what it proclaimed to be “every word in the english language” (in reality, the creator used a word list he found online [now lost] which had no basis in any dictionary), and was followed with much interest and engagement. Several great articles: Washington Post, Guardian, Gawker. Inspired many spin-offs such as @nondenotative (every non-word), @everywordisgay and @fuckeveryword – themselves creative interpretations of the vast quantities of information we have available for our use.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s