Journey to Democracia En Red
As a politically motivated person turned programmer by Computer Science 1 and 10, I explored the prospects for large scale online decision-making and discussion and direct democratic governance.
Many questions arose. What features should such a system have? Who would use it? How could it replace current political and organizational decision-making? How could it be secured? Is it desirable? Although this story ends with an ISTS funded trip to the Democracia en Red institute in Argentina, it begins with an obsessed College student researching radical democracy.
Part 1: Background Research
I started reading, searching for past and current methods, frameworks, and experiments. The book Online Deliberation: Design, Research, and Practice was a good start. A collection of papers compiled by Todd Davies and Seeta Pena Gangadharan, it introduced me to many of the experiments, challenges, and perspectives in the design and evaluation of online discussion and voting systems. Three papers were the basis for one feature I’d propose to DemocracyOS.
The first, “Decision structure: A new approach to three problems in deliberation” by Raymond Pingree introduced the idea of structured discussion, where each comment had an explicit relationship to the comment it was posted as a reply to. Pingree developed a system targeted at problem solving, where users could post problems, causes to problems, solutions to problems, etc.
The second, “PerlNomic: Rule Making and Enforcement in Digital Shared Spaces” by Mark Phair described a game with self-modifying rules. Using the programming language Perl, the program allowed users to vote on changes to the program. This introduced the concept of a user-modified, evolving system in the context of online deliberation.
The third, “Design Requirements of Argument Mapping Software for Teaching Deliberation” by Matthew Easterday, described the aesthetic and functional requirements for argument mapping software, designed to display and externalize the logical relationship between statements.
From the fusion of these three ideas, I developed the idea of a visual structured comment system that was configurable (the relationships between comments can be modified by the user).
Next, I explored theoretical questions — what would the ideal system be? On what ethical grounds should one compare systems? How would one transition to such a system? Since it felt odd to draw inspiration for a direct democratic system from political theory that did not prescribe it (Locke, Hobbs, American founding fathers, etc.), I read the book Participatory Democracy: Prospects for Democratizing Democracy, a collection of essays compiled in the late 1960’s and updated again in 2003.
It had a distinctly utopian feel to it, written by a group of people who saw empathy, play, symbiosis, and mutual aid at the core of human organizational participation.
Their angle on democracy was specifically and unabashedly socialist, viewing the democratic ideal as the democratic control of everything, including economic decisions at every scale. As a result, the authors prescribed democratic principles for all economic and social organization, not just limited governance.
Crucially they also related the two, arguing that a population made to feel continually powerless and deskilled in their day to day work would not feel powerful and engaged in governance. The inverse was that creating small scale democratic decision making bodies alongside existing hierarchical ones would create the experience of successful decision making necessary for confidence in its large scale application.
Thus, the participatory democratic theory of change is to create an alternate social and cultural reality — a network of democratic organizations built alongside the existing ones — that could be progressively built up to finally offer a competing framework for social organization. This “dual power” strategy, although devised before the Internet, seemed like the Internet’s calling.
This lead me to choose DemocracyOS over Loomio, Liquid Feedback, and the various other open source voting platforms. DemocracyOS was the only one that saw democracy as something that both governments and organizations do, and the two as fundamentally related.
Incidentally, the choice of DemocracyOS took me to Argentina, where it had been founded to build the software that would power the newfound Argentinian Net Party (similar to the European Pirate Parties), that sought to inject direct democracy into parliamentary procedure by electing “trojan horse” candidates that would vote based on the consensus of the user’s on the site.
The location of DemocracyOS in Argentina brought up another coincidence: the third component of my inspiration came from an almost-was democratic experiment in neighboring Chile.
Brain of the Firm
Although the idea of of using electronic tools to facilitate closer and more direct popular influence may now appear innovative and revolutionary, that is only in our context of our world of dysfunctional pseudo-democracies. One must imagine that any truly participatory democratic society, as it technologically progressed, would naturally update its decision making methods to use the latest communication technology.
However, you don’t have to only imagine — this happened when Chile, in the early 70’s, elected a democratic socialist president, Salvador Allende, marking the first time a real Marxist had been democratically elected. Allende moved quickly — although a strict constitutionalist committed to act only within his legal power and democratic mandate — he exploited loopholes in government bureaucratic power to aggressively nationalize the major industries: steel, mining, and fishing. Immediately after, he raised the wages of the workers by some 40%, drastically raising food demand and therefore prices, creating an trade deficit because of increased food imports. Chile therefore needed to improve productivity in their domestic export-oriented industries while maintaining the low level worker control desired by Chilean socialists, so they brought in Stafford Beer.
Stafford Beer was a fiercely anti-ideological British management consultant, who had invented the discipline of management cybernetics: the application of cybernetics, the generalized study of command, control, and regulation, to the management of industrial enterprise. Beer viewed human organization in terms of variety: the number of possible states of a system (closely related to information theory’s concept of entropy), and was interested in how human organization reduced variety to aid decision making on a large scale.
This was the inspiration for platforms — a system where opinions on multiple topics could be grouped together. This would allow a reduction in variety for the end user. Voting with a platform would be optional, to allow each user to decide how much of the complexity they would like abstracted from them.
Part 2: DemocracyOS
I arrived in Argentina the morning of September 15th disoriented and out of it. I had slept the 12 hours of the flight, but that was it. Since my hostel didn’t have check in until 5pm, I went to work.
That sleepy and dull version of myself then explained my ideas which would become the basis for my work for the next 10 weeks, and I encountered for the first time the language barrier. Guido, the CTO, was technically right — everyone spoke English, but that doesn’t mean communication was seamless. Diagrams were very useful. The development team didn’t have anything in particular for me to do, so they told me to get started on whatever I wanted to. I began working on platforms.
A few days later, I met with the political director over coffee, and gave him the pitch. This included the variable comment system, platforms, and the overall cybernetic vision of democracy and social organization. This time I didn’t have paper to draw diagrams as I spoke, so I used hand gestures and salt shakers to create a visual space on the table. I could tell he was personally both impressed and excited, and he told me that I would present all of this to the office next Tuesday.
That Tuesday was the most nervous I had been in a while — although half of the room was people who had already heard what I was saying, the other half was difficult to reach. Although I’ve done a good amount of public speaking and presentation, so many things — the language barrier, the personal investment in the material — made this different. The ultimate outcome of the meeting was, “Hm. Worth a try. Go ahead and do it.” I had my work cut out for me for the next 10 weeks.
After that, I got into a routine. I would arrive at work at around 10, program until 12:30, eat lunch until 2, program until 5:30, and play ping pong until 6:30. On Tuesdays there were office meetings, Fridays we went to eat Schwarma, and on the weekends I hung out with my co-workers or some other American students I met while there.
Everyday we would drink yerba maté — a highly caffeinated Argentinian herbal tea drunken out of a gourd through a spoon-straw. It was a communal drink, so the person who prepared the maté would pour a little bit of hot water in the gourd for the next person, get it passed back, and repeat. By week 2, I could successfully prepare and facilitate maté drinking for the 4 programmers at my table. I still drink maté even after I’ve come home, but it’s frustrating to drink it alone.
By week 4, I could understand a lot of the Spanish I heard, but more or less gave up on trying to learn to speak it. There were simply too many words I did not know.
Around week 6, the team released version 1.0.0, which included multi-forum functionality. After this, there was a period of time where we learned the “lean model” of start-up development, and thought about future development of the product, asking fundamental questions like “Who is the customer?”, and “What are we trying to make?”. This included drafting tagline product description type statements, which gave me the opportunity to make what I think was my best joke for the 10 weeks I was there. Highlighting the conflict between many of the features I was proposing and the need to maintain a clean and simple user interface, I joking suggested “DemocracyOS: A Dead-Simple Voting Platform Unless Ben Gets His Way”.
Week 6 was also when I made the transition from working on platforms to the variable comment system. Although I finished a workable version of the feature, it still needed some more work to be really usable. That’s still on my to-do list.
Around week 9, things began to get bittersweet as the end quickly approached. I was finishing up a simple proof of concept of the comment system I was advocating, while recognizing that I after I left they would keep doing what they were doing, except without me.
On the last Friday at our weekly schwarma eating, I gave each of my co-workers a joke present (a stick, a map, etc.) with some supercilious reason it was actually an earnest gift and represented a unique personal bond. This was the most authentic display of gratitude my teenage allergy to sincerity would allow.
I left for the airport later that night, filled with a great appreciation of the time I had spent there. The first day I wrote that Argentina really felt like a foreign place, and I wondered whether over time it would get less foreign. As I drove away through parts of the city I had never visited, still hardly able to have a conversation with the cab driver, the answer was obvious.
Of course Argentina is still foreign, but I had made real friends and connections, and those were anything but.
Part 3: Aftermath
After arriving home and slacking my way through a period of time where I was supposed to have been polishing what I worked on there, I missed my friends but communicated with them less than I thought I would.
I learned quite a few things about the e-democracy project — mainly that a vision of how democracy could work was insufficient. You still needed users.
I’m quite certain that I’ll continue to work on radical electronic democracy through my academic and professional life. Maybe I’ll even start a branch of DemocracyOS in the United States. I can only hope that I some point I’ll have some excuse to go back.