[iDC] The problem with social protocols

Elijah Saxon elijah at ucsc.edu
Sun May 23 08:17:04 UTC 2010

It is nice to see the growing wave of interest in 'the facebook
problem'. In a nutshell, the task at hand is to figure out how to
transform social networking communication into an agnostic protocol like
email, rather than a set of set of protocols that serve facebook.

The debate around these issues focuses on the question of 'open
protocols'. Facebook has open protocols, like "open graph", which are
really designed to expand Facebook's powers of surveillance and their
ability to commodify user behavior. Google has open protocols, like the
ones used with Buzz, that are designed to wrest control of social data
from Facebook, and cut Gooogle in on some of the profit to be made by
the detailed surveillance of social networks.

In the free software world, the debate is also about which open
protocols to use (or invent) to build a set of 'social protocols'. One
of the unresolved tensions in this debate is how secure the protocols
need to be. Many people hold the belief that liberation from the 'data
silo' of Facebook is good enough. From my perspective, it is unfortunate
that much of this debate carries with it the ideological baggage of
'empowerment' that has been central to the 'web 2.0' project and is
deployed frequently by Google and Facebook.

The rhetoric of empowerment is worse than co-opted--it has been suspect
from the beginning. If there is a foundational flaw in the discourse
around Web 2.0 it is that talk of empowerment hides the fault lines of
power. This allows Google to write about how they are saving the world
by embracing open protocols (they literally say that), while
conveniently ignoring the fact that the protocols they are embracing are
designed to allow Google maximum access to the monitization of user's
desires and social relationships.

First and foremost, an acceptably secure social protocol would prevent
service providers you do not trust, but still need to communicate with,
from building a detailed graph of your social network.

We need to start thinking of the social network as both intimate data,
and also highly political data. We know far too much about the power of
social network analysis to treat the social graph as anything less. When
someone's sexuality can be predicted with a high degree of accuracy by
social network analysis alone, or when there are mathematically
efficient methods of disrupting a social movement by breaking the least
number of network connections, it is time to start protecting our social

I have been following closely the debates in the free software world
around the 'facebook problem', and I am a bit worried. What is needed, I
believe, is a recognition of how 'open protocols' are used as weapons by
Google, Facebook, et al. Hopefully, we will start taking seriously the
significance of our social graphs and the various ways they can be
exploited for social control (to say nothing of constitutive forms of
power, like the emergence of surveillance-directed subjectivities and
the problem of technologies of the online-self).

Some 'social protocol' proposals do take the social graph seriously, and
perhaps they will be adopted... once all the others fail. :)

Viva La Protocol!


Elijah Saxon
Sociology PhD Student
University of California, Santa Cruz

More information about the iDC mailing list