In Praise of the Ephemeral

There is a creeping view in our society that data is forever. For some reason whilst we are happy for our spoken words to vanish into the ether, our written words and media must be hewn in stone (or at least backed up at an off-site data centre). Judge of the Nations, spare us yet, Lest we forget — lest we forget!

The art of Andres Amador is beautifully fleeting

This week ECS is hosting a series of Web Science Industry Seminars, and I was fortunate to attend two presentations that made me think about our society’s attitudes to the preservation of digital information, and wonder that maybe we have become distracted by the heft of forever data.

Richard Boulderstone from the British Library described their continuing effort to record the publications of the United Kingdom, which not only raises the issue of how you guarantee the future integrity of digital holdings, but also how you start to archive the massive and growing data avalanche that is the UK’s web-presence. Natasa Milic-Frayling from Microsoft then picked up the baton and described the technical challenges of archiving web materials.

Natasa’s central point was that a digital artefact as we see it is in fact a projection of some underlying digital file, and that to archive the artefact we need to capture both the file and the method of projection. It’s a nice analogy of the problem, and of course there is lots of complexity in capturing the method of projection, especially in an interactive and distributed medium like the web where a single page can have hundreds of dependancies and variables that affect its presentation.

Both Richard and Natasa were describing real and valuable challenges, but I was reminded of Viktor Mayer-Schönberger’s idea of computers that forget, and struck by how the need to record everything seems to be a growing social imperative, as if our culture is clutching in panic at dropped change that is bouncing and rolling through its fingers.

After all, total preservation is not something that we have historically managed or even tried to achieve. An artefact is a projection of a file in the same way that a play is a performance of a script, but we don’t feel the need to record every performance of a play – in fact their very transience is seen as special – and where do we draw the line? What about the other subtler aspects of our experience (who we were sitting with, what the weather was like on our way to the theatre, what we said as the curtain fell)?

You could argue that we are only now achieving the ability to record everything, and that this will improve and enlighten our future selves, but I think that we would be losing something in such a world. If data is ubiquitous then it becomes valueless; perhaps some data needs to be ephemeral rather than recorded and made weighty with time.

None of this is really new – certainly not in terms of data protection and privacy – but now the issue seems to be affecting our daily use of technology. For example, in recent weeks this obsession with the permanent record has crept into our national conversation about social media. The false accusations made against Lord McApline (not to mention his robust response!) has marked a shift in tone in the way we discuss Twitter. What was initially seen as an extension to our conversational habits, deliberately light and inconsequential (tweet, tweet!), seems to have become something viewed as a full-bodied publication with all the responsibilities and risks that this implies.

While I agree that any act of speech carries with it some responsibility, I dont see why we cant differentiate between different speech contexts, and that may mean accepting or even embracing the ephemeral. When we discuss digital literacy we include aspects such as information literacy, where readers consider the provenance of data, and perhaps we should consider the intention of data too. The responsibility on how to interpret something said is as much on the listener as it is on the speaker.

So perhaps we need to make space in our digital lives for the ephemeral. Not to manage the technical challenges of remembering but because it is valuable and healthy for a society to forget. Not all data need be forever.