Friday, August 17, 2012

Creating a Modern Feudal Order

'Plight of the Data Serf'
"The seventeenth century was a period when old answers were inadequate, but new ones had not yet been found to take their place," wrote James Billington in his 1966 book, The Icon and the Axe: An Interpretive History of Russian Culture.  Describing the historical transition of medieval Russia into the modern period he added, "The inevitable waning of old Muscovy could well be described under the first three chapter headings of Johann Huizenga's classic Waning of the Middle Ages: The Violent Tenor of Life, Pessimism and the Ideal of the Sublime Life, and The Hierarchal Conception of Society." [119]  Billington had good cause to promote these three chapter headings as pithy summaries of the period in Russian history.  The ascension of the Romanov dynasty in 1613 brought radical change to the structure and operation of the Russian state.  Not only did serfdom become legal through enactment of Russia's first printed law code, the Ulozhenie of 1649, but in doing so the newly established Romanov dynasty ensured its nigh-absolutist rule through the cooperation of the nobility.

In late June of this year,  the Twitter Developer Blog posted an announcement titled 'Delivering a consistent Twitter experience'.  The overall theme of the post, consistency, took on worrying overtones by the end as developers read "in the coming weeks, we (Twitter) will be introducing stricter guidelines around how the Twitter API is used."  This announcement generated considerable anxiety among developers because the API is the primary protocol applications utilize in order to interact with Twitter.  Stricter guidelines were generally understood to mean that the owners of Twitter wanted to assume tighter, centralized control over the platform.  Then, yesterday, those stricter guidelines arrived.  In a new post on the Twitter Development Blog, 'Changes Coming in Version 1.1 of the Twitter API', Michael Sippey (Director of Consumer Product and author of the previous Twitter announcement listed above) spelled out upcoming changes being made to how third-party developers can access and utilize the API.  Display guidelines for third party clients are now becoming display *requirements*.  The amount of users allowed to access the Twitter API through third-party clients will be limited, dramatically.  Marco Arment, developer of the very popular Instapaper app, provides a great breakdown of the impending changes on his blog.

There would appear to be little in common between the situations described above- but deeper inspection reveals that serfdom, and the process behind its implementation, unites both.  Russian serfdom of the seventeenth century bound peasants to the land in order to reliably extract economic resources and appease the nobility in exchange for their allegiance and loyalty.  Data serfdom of the twenty-first century binds users to centralized platforms in order to reliably extract verifiable data and appease marketers in exchange for cash and control.  The story of how Russian serfdom came to be finalized in the Ulozhenie of 1649 shares similar parallels to how Twitter and other data platforms seek to create a 'Hierarchal Conception of Society' through access to their data fiefdoms.  Twitter's API announcement, like those made by other data lords, is just one step towards realizing a return to the feudal system in the age of data platforms.

Serfdom, like any complex and normative legal system, does not simply come about by happenstance or decree.  It is the product of several decisions that accumulate into a binding corpus.  Russian Law Codes of 1497 and 1550 initially allowed peasants to leave their landowners estates, but only during a two-week period that occurred at the end of the agricultural season in November.  Between 1550 and 1649, the Russian state continued to restrict the right of peasants to move, largely through backing gentry requests to have escaped serfs returned to their owner's estate.  In fulfilling these requests, the Russian state not only improved its stature among the various landed lords but also ensured that economic resources- in the form of military readiness, agricultural production and internal population migration- remained stable.  This was a very compelling economic arrangement for a state situated with an abundance of natural resources yet few, and often flighty, available people to work the land.

If we take a similar survey of Twitter's demesne, some interesting factors begin to align.  Back in 2006 when Twitter first debuted, many standards of form and protocol were not yet established.  Common features used today like the retweet, the @ reply, and even hashtags were not embedded in Twitter's design initially and came about only through the formation and later adoption by users of these standard conventions on the platform.  Outside developers began to craft their own Twitter clients (think Tweetdeck, now owned by Twitter, or Tweetbot) thanks to the generous opening of Twitter's API, which allowed for both reading of the Twitter stream (the ability to bring your stream to a client) and writing to the stream (the ability to post a tweet from Tweetbot, or another client).  To put it in feudal terms, the Twitter lords staked out a new data domain and actively recruited both laborers (average Twitter users) and lords (third-party developers) to settle so that the 'land' could be developed and put to greater informational/economic use.  While these third-party clients allowed new developer lords to effectively create their own titles and dynastic lines, Twitter could abide this fragmentation of power because it brought more data serfs into their collective domain.  API access facilitated the creation of a new feudal order.

This last point is important, as Twitter actively sought two levels of users- one level as common laborer and another as a sort of landed gentry to help manage/recruit common laborers.  Developers, given a large degree of access to the data domain of Twitter via the API, could become petty data lords themselves, creating applications that grew Twitter's data domain and facilitated collection of user data crops from regular serfs toiling away in multiple 140 character plots.  This move allowed Twitter to benefit from other's labor in promoting their service and, as a natural consequence, gave these petty data lords a compelling reason to make their applications the best in order to secure the largest possible data crop.

The Russian state, too, promoted such a distinction between 'users' and 'petty lords' through the pre-serfdom institution known as pomeste.  Cavalrymen, the distinct element that helped Russia acquire new territories through warfare during the 15th century, were given plots of land that contained settled workers at the conclusion of hostilities.  With this plot of land and supplied work force, the cavalrymen could not only support themselves and the maintenance of their equipment (horses, weapons and armor were all very expensive) but also provide low-level stewardship for a, then, growing Russian state.  Success of the pomeste system provided one of the building blocks for Russia's eventual embrace of serfdom, and it could be equally argued that the success of Twitter's early API allowance for developer uses provided the building blocks for the, now, apparent regulation of the Twitter experience by the platform's owners.

Under the cover of 'consistency', Twitter is restructuring the rules governing their own petty data lords behavior and acquisition of power through collection of data serf production.  Last year, developers were warned to not build "client apps that mimic or reproduce the mainstream Twitter consumer client experience', i.e. to not develop alternative visions of the Twitter interface.  Now the company is sending signals that the 'consistent' Twitter experience will mean less developer involvement in the running of the platform.  Instead, developers are encouraged to put their efforts towards the new 'cards' feature, whereby a tweet acts as a container of shared content like a YouTube video or photo from Instagram.  Yet this is far from the relationship Twitter initially cultivated with developers and signifies a sea-change in how the company plans on managing its growing count of data serfs.

Not only has Twitter effectively declared its demesne to be that of the *entire* Twitterverse, but also that its vassal developers, once allowed to carve out their own estates in the Twitter kingdom, now must make due with much smaller plots of land, i.e. 'cards', limited data serfs to work those plots, and loss of title.  Enter  The new upstart platform raised just over $800,000 to back its vision of a Twitter-esque experience without the need to resort to advertisers for support.  Instead of providing a free service, asks up-front for users to pay a $50 subscriber fee for one-year access.  (This could change- the venture is still so new, many things are changing on a daily basis)  Developers pay more, but they are given access to's API and the ability to build any application they desire to work on top of the platform.  

If this sounds familiar, it should be- this is exactly the same process Twitter pursued in its early days.  Yet the idealism behind backed to a large degree by the insistence on a subscriber model- to remain free from advertising puts an interesting spin on the feudalization effort currently pursued by large data platforms.  Whereas the goal of the data lords is to continuously and reliably extract verifiable data crops for use in marketing, looks to promote what appears to be a 'Guild' like system.  Instead of data fields, seeks to cultivate guild members of various rank whose expertise is the only limiting factor for advancement on the platform.  Playing with the feudal metaphor,'s 'Guild' platform can be seen as a very urban response to the growing power of agrarian data lords.  Of course, is still very much new on the scene and it has yet to be demonstrated if this alternate vision can survive and thrive among the presence of feudalistic data lords, or if it will become a mildly differentiated data lord itself.

However, even the appearance of this new 'Guild' type of data platform will do little to alleviate the plight of data serfs around the Internet.  As the recent moves by Twitter demonstrate, there is an increasing trend among data lords to isolate their kingdoms from each other as smaller, yet powerful, vassals like Instagram, Spotify, and Tweetdeck begin to declare their allegiances to specific platforms.  The feudalization of data platforms is in full swing, leaving data serfs to suffer the brunt of their liege lord's designs.  "The seventeenth century was a period when old answers were inadequate, but new ones had not yet been found to take their place"- the same could be said of today.