Tuesday, June 19, 2012

From Data Self to Data Serf

Photo by L D M
"We belong to you, but the land belongs to us." - Russian Peasant Refrain

"But even when I am at a loss to define the essence of freedom
I know full well the meaning of captivity." - Adam Zagajewski, 'Freedom'

"You can't be what you were
So you better start being just what you are
You can't be what you were
Time is now and it's running out…" - Fugazi, 'Bad Mouth'

A few months ago, in preparation for my (then) upcoming 'Theorizing the Web' presentation, I tweeted the peasant refrain quoted above and made a comment to the effect that this is how, I believe, many feel when it comes to social media platforms and the data those platforms contain.  Essentially, the modern phrase would be 'We belong to your platform, but the data belongs to us."  Upon further reflection, however, I've started to think that this isn't the case at all.  Take, for example, the difference of opinion between American data platform owners and privacy advocates from the EU or the recent piece by Alexis Madrigal on the Atlantic titled The Perfect Technocracy: Facebook's Attempt to Create Good Government for 900 Million People.  Looking at the social media's sites process for regulating content on its data platform Madrigal states, "People know that Facebook controls a large slice of their digital lives, but they don't have a sense of digital citizenship."  I think the problem goes beyond the ideal of citizenship, even though Madrigal's statement hints at the larger issues.  Evidence grows that we are becoming serfs amidst the growing fields of cultivated data- and there has yet to emerge an articulation on the relation of our serfdom to data platforms like that embodied by the peasant's refrain.  Why is this so?

From my perspective, the heart of data serfdom centers on a question over what counts as certified, verifiable knowledge and the degree to which that knowledge is permitted to circulate or be modified.  Data platforms embrace stable declarations of the self through updates, likes, pins and other markers that, once declared, are difficult to change in content or meaning.  Even if one erases, or 'unlikes', something, the impression generated by the data platform of your data self remains tied with this now etherial 'like'.  Because we cannot change the content of these declarations easily, the validity of the message becomes even more key, even more vital.  This presents some unique challenges with regards to the relation of our data to ourselves.

Low-Mobility Value Facebook's 'Like' Button Produces
Photo by Eric Schwartzman
I would like to investigate the writings of Rob Horning (@marginalutility) and his focus on the 'data self', which explores how neoliberal capitalistic practices capture users in a sticky web of commodity production/exploitation.  His perspective and analysis helps clarify points I would like to make regarding how data platforms current utilization of low mobility knowledge constructs fits nicely with the evolution of documentary regimes into their new hybridized form.  Whereas pervasive documentation once allowed governments and others in power to shape the normative discourse (or at least put it into quantifiable terms), the new data platforms allow those in power to not only continue pervasive documentary practices but also avail themselves of that practices ills through integration of highly personal intermediation.  Social media is the embodiment of this intermediation.

In Advertising and the Health of the Internet, Horning addresses Atlantic writer Alexis Madrigal's piece detailing numerous companies that exist solely to track users data.  Picking apart Madrigal's premise behind the inherent value of this data, Horning states that the natural extension of Madrigal's logic would have one conclude that, "the health of the internet…depends on the degree to which we can turn thought into marketing through the process of circulating it."  For me, circulation is the key term in this statement.  Why?  Because I believe a key insight for understanding the 'data self' centers on the operation and interaction between high mobility and low mobility knowledge constructs, terms I've created to deal specifically with the circulation and adaptation of information.  In order to understand the terms of my critique, it would be best to take moment and define the terms and their execution as I intend.

Data platforms embrace, wholly, the configuration of low-mobility constructs, especially when it comes to articulating a data self that not only provides panoptic power but also algorithmic control in how we view our relation to others and ourselves.  Low-mobility constructs, generally, aren't modified through their transmission or reception.  I've often brought up the book as a prime example, but the same holds true for a Facebook update or Amazon purchase or Pintrest post.  These online statements are essentially one-offs that everyone can see- yet no one can change your thoughts, or purchase, or 'pin' through their own transmission or addendum.  All they can do is create a low-mobility response, adding shades of inflection to a conversation that is both asynchronous and bound in stasis bubbles.

This in itself would not be revelatory.  But if we look at Horning's line stated above through the lens of low mobility a key point does arise.  Circulation of thought in low-mobility form is entirely dependent on validation.  Because low-mobility constructs are highly resistant to being changed through transmission, the very act of circulating these constructs hinges on validation by the transmitter and, to some degree, the recipient as well.  High-mobility knowledge constructs, in comparison, can enjoy the added value validation brings, yet their impetus for transmission rests solely on the inherent capacity to be changed by both the transmitter and recipient through circulation.  Its very mutability is what ensures its survival and circulatory value.  In the past, I've used rumors as a prime example of high-mobility artifacts and specifically examined how Russian peasants often successfully used rumors to obfuscate demands made by Imperial authorities exemplified in low-mobility edicts or laws.  (See, for example, my Theorizing the Web presentation paper discussing this phenomenon in more detail)
Data platforms and the 'data selves' they produce, in order to generate some sort of commercial value, embrace low-mobility methods of knowledge production because the incessant need for validation as a precondition of circulation not only makes our various crops of data stable, (hence Target can accurately predict pregnancy-changing consumer behavior before even you are aware of it) but it also traps how we view ourselves in purely low-mobility, unchanging terms.  The only change allowed is that validated by the platforms, whose certification of knowledge creates a self-fulfilling and continuous loop of data.  This, in turn, creates a new view of the self whereby the past is illusionary and the future is now.  Meanwhile, the present languishes amidst declarations based both in the past (I 'liked' this post, I 'pinned' this item) and situated in the future projection of the self (think saving articles for 'Instapaper').  It should also be noted that this view of the self, largely predicated through digital terms, impacts the operation of augmented reality the self inhabits.  (An effect Nathan Jurgenson calls the 'Facebook Eye')   This is the kernel of data platform enserfment so many find themselves in today.

The 'landed' lords of the platforms, and the various petty 'landless' lords they sell our 'cultivated data' to, depend on this stable projection of the self trapped forever in the present.  If we could change our thoughts and our projections on the fly, carefully arranged cultivation processes would be tossed into a state of near-anarchy.  The asynchronous effects demarcating our lived self from the data self would grow to such proportions that any attempt to place a commodity effect on our production would yield false, shadow selves and thus become worthless to data platforms and marketers alike.  Moving targets are much more difficult to hit, especially when your gun is fixed in place.

Photo by Michael Mandiberg
If we look at the history of information through the lens of mobility, a progressive shift from the predominance of high mobility constructs, embodied by oral conceptions and utilizations, to that of low mobility constructs, introduced first by manuscript culture but later reinforced by print culture, can be detected.  Our current digital platforms of data only reinforce this trend, as I have tried to explain above, binding our actions increasingly in terms that are indicative of low mobility expressions and shunning forms that could allow the possibility of high mobility transmogrification.  In doing so, the new data platforms managed to solve an age old problem that plagued panoptic potential of documentary regimes.  By channeling the documentary process through a highly mobile (but not high mobility) process of algorithmic control, platforms can enjoy the benefits of both documentation and control with only minimal disruptive potential.  Increasingly the self becomes merged with the projected data self and the asynchronous friction that this process would typically engender becomes smoothed over due to the ease by which the data self can be fed, nourished and anesthetized.  Numb to the pain, we no longer recognize that our data is not our own.  No need arises for high mobility responses to low mobility control.  We become data serfs with few means to shape the terms of our bondage.

Russian peasants at least possessed access to the high mobility world of oral reinterpretations and rumors to challenge low mobility edicts or laws judged 'unfair' or 'unjust'.  The current generation of data serfs putter along without these tools or techniques, in effect declaring through their sustained cultivation of the data self the idea that both our data and ourselves belong to the platforms.  This is a dangerous position to be in, something Horning notes towards the end of Advertising and the Health of the Internet when he states,
"If our public lives are underwritten by our value to advertisers, our public selves will end up indexed to that value for everyone, and our private sense of ourselves will be to a degree dictated by the boundaries of the sensorium marketers can create around us with increasing specificity."
In Social Graph vs. Social Class, Horning follows his 'boundaries of the sensorium' concept to its natural extension by examining how Facebook's 'Social Graph' articulates both social organization norms enforced by social media and how those norms reinforce class-based analyses that are supposedly absent in the data platforms construction/execution.  Many people understand the 'networked' concept of social media platforms, yet fewer probably view their connections, in Horning's words, in terms of 'lines of competition as well as mere affiliation."  Asymmetrical power relations are inherent in data platforms just as they are in life.  Yet while many can accurately and articulately define those asymmetrical relationships in their personal, face to face interactions, they more easily gloss over the subtleties of such relations when using data platforms.  

"The social graph purports to passively record social arrangements that emerge organically and thus reflect some sort of true and undistorted account of how society works," states Horning, yet I would contest this point by saying that data accrued through cultivation of low-mobility sources cannot be organic- it is anything but organic.  Real communities, ones that establish possibilities for membership to make meaningful changes, exist on a mix of low and high mobility constructs.  They are chaotic, changing entities.  The lords of the data platforms, knowing the pains previous landed lords went through, sought circumvention of this community aspect through their algorithmically defined user interface.  Horning believes that social media makes identity construction fluid and able to be revised in real time- I believe it creates identity stasis as the constant accumulation of low-mobility statements continually weighs down the lived self with the constraints of the data self.

But how do we escape this serfdom?  Russian peasants were largely unsuccessful in their quest to achieve freedom, so why hold them up as paragons for a 'second' serfdom now occurring with the rise of data platforms?

The answer to both lies in finding ways to integrate high-mobility potentials into the data platforms on which our data resides.  While not a perfect example, Twitter has shown how this can be accomplished.  The retweet and hashtag features did not initially debut with the system- instead they evolved through user-centered invention and innovation, becoming an official part of the Twitter UI only in mid-2009.  These were not strictly high-mobility constructs shining through, rather they represented the exploration of potential space the Twitter platform allowed through the boundary-pushing behavior of users not engaged in a tightly defined algorithmic experience.  The inclusion of oral testimony as a primary source for use in Wikipedia articles, something I've written about, represents another potential mixing of high and low mobility sources, though not without some compromise.  Just as Russian peasants challenged edicts and textual dualist norms espoused by Tsarist authorities through use of 'augmented' oral constructs, so too do we need to find methods by which the low-mobility nature of data platforms have to contend with high-mobility critiques generated by an authentic community of users.  

It goes beyond voting on potential privacy rules on Facebook- this means putting the actual creation of privacy regulations in the hands of the users themselves.  Of course, this would mean threatening the stability of data reaped by data platform lords, not to mention an acknowledgement of a general consciousness on the value of the self to the process of data gathering.  If we fail to do so, we risk becoming serfs on a scale greater than that imagined by the most ardent Russian landowner.


  1. This comment has been removed by the author.

  2. what do you mean by: " the value of the self to the process of data gathering."?

    1. This would be an acknowledgement by the users that their personal data- often construed by the data platforms as a glimpse into your 'self' (tied to sense of identity)- has intrinsic value and that any use of such data should be in the hands of the users. Now this is, admittedly, a utopic notion, however there is little evidence that users even demand a fraction of the control this ideal would hold high. And I think one factor for this apparent 'sedation' is that these platforms act like they offer you a service that you, yourself, cannot divine- that being the composition of the self. Of course, I (and many others) would challenge this notion, but nevertheless it gets to the appeal and 'logic' of the data platforms position.