Decentralizing Data (and Publishing)

This past week I’ve been thinking more and more about issues of centralization and decentralization in terms of digital infrastructure. My thinking about these things are inchoate. Obviously, some of it stems from conversations about physical nodes in networks such as data centers, anxieties about the corporate organization of our online and interconnected digital world, and the role of institutions in mediating the dissemination of knowledge and information.

Eric and Sarah Kansa’s recent article, “Open Context in a Changing Context: Data Publishing, Interoperability and Governance” in Internet Archaeology, prompted me to think about the relationship between institutions and knowledge making in archaeology. The article is good and worth reading for all sorts of insights into how Open Context works. One point that stood out to me, however, was that unlike many other data archiving and publishing services (and I deliberately conflate the two even as I understand the former requires more substantial infrastructure than the latter), Open Context does not have an affiliation with a pre-existing institution such as a federal agency or a university. The Kansas argue that this reflects the capital intensive requirements of long term data preservation (which, incidentally, parallels the development of libraries), but also represents a vulnerability especially for those repositories the require federal or state funds for maintenance. Open Context, in contrast, relies on a range of different repositories as backstops for its data publishing with an eye toward ensuring that even amid the increasing vagaries of institutional commitments and priorities, the publish data remains persistent. 

(By the way, you should go and read this article on its own merits which go far beyond my slightly incomprehensible rambling here. I’ve had the pleasure of working with Eric and Sarah Kansa on and off for most of the 20-odd years that Open Context has been thing. I’ve also published data with them from the Eastern Korinthia Archaeological Survey and our two projects at Pyla–Koutsopetria on Cyprus). 

This flexible and recentered approach to data publishing (rather than the tendency toward centralized data archiving) parallels their broader approach to data. Rather than imposing a standardize template on data that Open Context publishes, they publish archaeological data according the schemes that the project itself defines and presents. This means that comparisons across datasets published by different projects is more complicated and requires us to manage fuzzy alignments to produce meaning. This is a slow process in that it mitigates against the efficiencies promoted by standardized schemas. It also divests itself of any authority inherent a centralized schema. 

I’ve been thinking about this stuff against some recent works on digital infrastructure. Britt Paris’s recent book, Radical Infrastructure, which I’ve only read part of, considers the limits and challenges of our digital infrastructure and its shadowy collusion with the interests of the state and capital. Of course, these interests may not align with the interests of the users and, in fact, may run strongly counter to users political, economic, or social commitments.

Paris sites work like David Nemer’s Technology of the Oppressed(2022), which considers the ways that people living in Brazil’s favela communities created networks using improvised methods. I’ve only skimmed this book so far, but it clearly is something that I need to read more carefully.

The point of my post today is to start to think a bit about how institutions mediate the interface between physical infrastructure (such as servers, cables, switches and so on) and intellectual infrastructure (ontologies, data structures, and other tools). Outfits like Open Context are interesting because they demonstrate how standing even slightly outside of these institutions gives them a position where they can offer a distinct perspective (and an implicit critique) on their operations. 

This is meaningful to me, in part, because I’ve been thinking about how publishing can work in this way as well. Presses (cough… like The Digital Press at the University of North Dakota) that operate on the margins of institutions (as I’ve argued before, I like to think of my press as cultivating the undercommons) or completely free from connections may occupy positions of critical importance. In particular, I wonder whether scholarly and learned societies could similarly work to support practices and structures that offer critical resistance to standardization imposed by capital, efficiency, and, state authority. Of course, I understand that we live in a deeply interconnected world and learned societies are not more free from the entanglements of capital, the state, and the demands for efficiency and standards. That said, Open Context offers us a glimpse of how even within systems that are inhospitable to the kind of engaged, critical, and even inefficient way of presenting data and building knowledge, distinct and creative ways forward are possible.