Walled gardens or gushing brooks? Public and private data online

“For me there are no answers, only questions” wrote P. L. Travers, “and I am grateful that the questions go on and on”. For each of the ethical pressure points of social media research - harm, copyright, data protection – we are uncomfortably close to Pamela Travers’ fantasy. 

One of the sorest pressure points of all is deciding what is private and what is public online. What data is sacrosanct and what is, as they say, ‘fair game’? A lot pivots on this question: when and how privacy is invaded, the importance and role of informed consent, the protection of identities and persona. In sum, the art of identifying and mitigating the harms inflicted by researchers tapping into vast corpora of social-digital data.

I can’t – and won’t – offer a clear-cut answer. But I can point to one of the reasons that the move towards answers has been so difficult, and that is the role of metaphor. Social media is a new social technology and new common habit. As so often happens when grappling with something new, we resort to a series of metaphors to relate social media to something older and more understood. Three powerful metaphors play a decisive role in conceptualising the private/public problem. 

Space – The Internet is a network of public thoroughfares, semi-public alleyways and cloistered ‘walled gardens’. Understanding privacy spatially in this way nudges certain factors into prominence, especially the local laws – or privacy settings - that govern a particular territory. 

Flow – The Internet is also understood as a network of rivers of information – with gushing rivers, smaller tributaries, brooks and lake. More prominent here are questions about expectations about where the information is going – how it is used, shared and disseminated.

Text – Finally, there is the idea of the Internet as a series of interconnecting texts, a library of libraries. Privacy in this sense implies the relationship between author and text, and especially the terms under which each text was published.

None of these metaphors are altogether wrong, but none are right. When you over-extend metaphors, you also inherit a great deal of unwanted baggage: false dichotomies, half-apt analogies, and blocked off possible solutions. These metaphors have framed our approach to social media research ethics and restricted our responses to them, in powerful, if often unseen, ways.

So a suggestion: the work that must now be done in social media ethics must recognise that the social media is a new thing, Joining a group on social media is not necessarily like joining an offline group. Publishing a tweet is not necessarily like publishing a book. And a discussion on a forum is not necessarily like a discussion down the pub. New concepts, new vocabulary and new approaches to research ethics – not metaphors – are needed now. Perhaps then, pace Travers, we’ll get some more answers.

Carl Miller is an Associate at Demos.