Computerization, Centralization, and Concentration

One of the most dangerous canards of the digital revolution is the one according to which distribution, decentralization, and democratization are the characteristic hallmarks of contemporary mass computerization.

To writers of earlier ages (Huxley, Orwell, Lem, Weizenbaum, Wiener, Mumford, Ellul, Roszak, just to name a few), such sentiments would seem shocking, because what they understood was the opposite: that computerization leads to–in fact, is usually specifically done to enable–the centralization and concentration of power.

I will leave reflection on how we got to this point for another day. Here, I simply want to lay out as plainly as possible the ways in which this is so.

To begin with: despite the way the rhetoric sounds, centralization and decentralization are not opposed, at least not in all circumstances, and the way to contest the thesis that computerization centralizes is not to produce examples of decentralization, but directly to confront the manifest examples of centralization.

Without belaboring detailed technical definitions of centralization or decentralization here–though such definitions are vital, and remain more necessary than many think (I would recommend Alexander Galloway’s discussion of decentralization in his classic Protocol: How Control Exists after Decentralization, and my discussions of concentration in The Cultural Logic of Computation)–my main point is about rhetoric, discourse, and ideology: that focusing on examples of computerized decentralization draws our attention away from the clear and obvious ways in which computers are being used to centralize and concentrate power.

What makes this hard to see at times is a series of confusions characteristic of our digital age. We tend to look at computing artifacts themselves, draw parallels between the hardware or software and the parts of the rest of the world they mimic or simulate, and then fail to remember to distinguish them in our thinking. Thus, the internet succeeds because it is literally a decentralized network, one that does not rely on a central hub or central computing server for its power. Yet the decentralized network specifically offers centralized and concentrated communication to its participants, compared to what existed before; that is its virtue. Human communication becomes relatively centralized and concentrated, even as the physical network is distributed. Similarly, we call “social networks” platforms that simulate social connectivity between human beings and institutions in software and digital networks, and then refer to them as if the software is the social network, when their efficacy is directly related to the underlying existence of the connections between human beings. We make metaphors for understanding, and then figure becomes ground, and vice versa.

Contributing to these confusions is the fact that, usually, centralization and decentralization are the flip sides of a single coin. By “distributing” or “decentralizing” computational resources across an environment, we create centralized nodes of power. Straightforward examples of this include RFID tagging, in which computational identifiers are distributed across a range of physical products (which looks like decentralization) but with the purpose of making more efficient supply chain management for the dominant, concentrated nodes in the network (such as Walmart and its distributors).

A related example has been uncovered recently in reporting on government surveillance of cell phone towers. Almost all writing and thoughts about cellphones treats them as the “mobile devices” they are; yet they (under current technology) require service nodes that may be every bit as centralized as early versions of telephone technology. Because they are called “mobile” and the devices are physically mobile, many of us wrongly assume that the technology itself is fully decentralized, but very much the opposite is the case.

Similarly, the decentralized internet itself depends on central nodes that are prime sites for State surveillance and interference. These network routers–nameservers and other central gates through which all “proper” TCP/IP traffic must pass–provide access points to the entire network that make much less clear the benefits to privacy one might expect from a truly distributed network. There is significant evidence–some might even say proof–that world governments use these central nodes to monitor and collect all internet traffic–all of it.

uranium enrichment centrifuge

We say that our media landscape has been democratized and transformed. Yet as Matthew Hindman’s masterful Myth of Digital Democracy (Princeton UP, 2008) shows convincingly, we in the US get our news from a smaller and more concentrated set of information providers, as a whole, than we did in the era of 3 big television networks (and thousands of local and regional print publications).

Google, Amazon, Facebook, and Twitter all thrive precisely because they centralize services within single platforms. In each case, they represent the digitization of real-world social phenomena that were far more distributed prior to their advent than after. It is precisely their nature as central, concentrated platforms that makes them powerful and useful. Wikipedia is a strong, centralized information source, having consolidated two or three slightly less centralized information sources (World Book, Encyclopedia Britannica) into one–but framed in the bizarre doublespeak rhetoric of distribution and decentralization. As those of us both suspicious and critical of the rhetoric of “Facebook Revolutions” and “Twitter Revolutions” have pointed out over and over, the visibility of actions and actors in these networks–often their most obvious “benefits,” but specifically of a “slacktivist” sort–provide much more centralized means for repressive governments (and other institutions) to track the actions of dissidents and whistle-blowers.

The US today is largely manged by powerful central actors, especially corporations, who do not shy away from computerization, as one might expect if the technology itself was decentralizing; to the contrary, corporations are among the chief users of and advocates for computerization. It is no accident that the digital revolution happens along with the move toward “big-box” or “category-cutter” retailers like Home Depot, Staples, and Walmart, the success of each of which depends directly on the extensive use of “innovative,” “new” digital tools.

These cases are among the clearest examples of “what computerization does,” and whether we think they are beneficial for individual consumers, there is little doubt that they have had clear and devastating effects on the “mom and pop stores” of “small-town America,” and that the empty parking lots of box stores make poor substitutes for the thriving Main Streets of decades ago.

It behooves us to try to keep in mind not just ourselves as users but every part of the social world that computerization affects. It might well be the case that the Home Depot customer has more “democratized” and less centralized access to more kinds of household tools and appliances than she had at the local hardware store. But it is hard to call that “democratization” or “decentralization” from the perspectives of either the owners of the hardware stores that closed, or even the towns that may well have wanted to maintain their Main Streets but were faced with no choice through the strongarm tactics of neoliberal capital.

A George Orwell or Aldous Huxley might think their nightmares have come true, watching us tweet about massive centralized resources like YouTube and Amazon under the name of decentralization and democracy.

Next: Centralization and the “Democratization” of Higher Education

This entry was posted in "social media", cyberlibertarianism, google, materality of computation, surveillance, we are building big brother, what are computers for and tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.


  1. Posted October 26, 2012 at 12:59 pm | Permalink

    […] The web is not very distributed or decentralized. There is a myth of digital democracy. When Amazon AWS has a hiccup, usually in its Virginia availability zone, half the American […]

  2. Posted November 9, 2012 at 12:59 pm | Permalink

    […] my previous post, “Computerization, Centralization, and Concentration,” I discussed how the fact that decentralization and distribution are genuine hallmarks of […]

One Trackback

Post a Comment

You must be logged in to post a comment.