Tools for Democracy / Distributed Journalism

Dan Gillmor points to an excellent example of distributed journalism in action over at Daily Kos. I was completely blown away by what I saw when I got there. I’m still trying to soak in all the background around the Plame Leak / Jeff Gannon thing but, to be honest, the specifics of this event are not as important to me as the general phenomenon occurring there.

This seems an obvious glimpse into a future where involvement by the general population in issues of import to the general population is increased substantially. Herewith some rant and analysis of our present toolset and humble suggestion for improvement.

If you believe, as I do, that the success of a society is directly proportional to the amount of active participation by its members, you may draw two conclusions given our present situation:

As a software developer, it is this second point that I’d like to concentrate on. After seeing what I saw on that Daily Kos thread, it became apparent that people need better tools. The amount of collaboration going on over there is amazing considering that almost everything is done using simple weblogs (diaries in Kosland) and lengthy comment threads. This system works fine for general discussion but the organization around this Jeff Gannon thing has some interesting traits that beg for more understanding from tools. It should be a testament to the will of the members of the Kos community that they are organizing so well using very basic tools. (Or maybe the simplicity is part of the reason it is working? Definitely something to consider.)

What’s going on?

Before I get ahead of myself, let me offer some analysis on the kinds of interaction we’re seeing in Jeff Gannon story and others like it for those who haven’t been following along.

Something happened a little while ago on Talking Points Memo that struck me as being significant in the same way as the Plame Leak / Jeff Gannon story is significant. Joshua Marshall was interested in finding out the names of the Republicans who were for and against the “DeLay Rule” (the new rule allowing Tom DeLay (R-TX) and future indictees to continue in their House leadership roles after being indicted). How reps voted was not public information and only Rep. Chris Shays of Connecticut came out to the press to say that he didn’t support the move. Shays mentioned that only a handful of other republicans voiced opposition, and so began the hunt for the Shays handful:

Do you work for a local newspaper or TV Station? Want an easy story? Call up the local Republican member of congress to see if they supported the DeLay Rule. Believe me, this one writes itself.

Rep. Chris Shays (R-CT) says a “handful” of members of the caucus opposed the rule. Let’s be extremely generous and say that’s 20 members. That still leaves 211 supporters. Who are they?

Not a journalist? Afraid you can’t play? Fuggetaboutit … You can play too.

Just pick a Republican member of Congress, call the number on their website and ask. Don’t be rude or confrontational. Just a simple question: Did Congressperson such-and-such support the DeLay Rule in the GOP caucus meeting on Wednesday. After you get your answer, drop us a line and let us know what you hear. Did they vote for the DeLay Rule or are they members of the Shays Handful?

In the days that followed, Joshua reported tons of responses from readers who called their representatives. Most reps claimed the vote was private (it wasn’t), or that they would receive a letter explaining their vote. A large number claimed that they were part of the Shays handful (more than a handful, in fact). Through the whole thing I think I saw only a single case where a rep admitted to voting for the DeLay rule.

But the specifics are unimportant. What I want to concentrate on is the general processes that made this and the Gannon story successful cases of distributed journalism / activism.

  1. Someone needs to drive. In the Shays handful the driver is Joshua. In the Gannon story, it’s SusanG. In both the cases, the driver has a better-than-average understanding of general politics, describes the initial issue, and (this is key) gives specific tasks that can be carried out by the community.

  2. The community comes back with answers.

  3. The community comes back with more leads.

  4. The driver is constantly integrating this feedback from the community into a central description of the issue. New leads are thrown back out to the community with direction on how they might best be pursued.

What’s Missing

What we need to concentrate on building tools around is better mechanisms for identifying, capturing, and tracking feedback and leads. I think the current infrastructure (weblogs, RSS, Technorati, trackback/pingback, etc.) is sufficient for a driver to get the initial word out on an issue. If it’s something people are interested in, links will put it on enough peoples radar so I think that will do fine for identifying issues initially. It’s the next steps that need work.

Joshua Marshall relied on email as the feedback mechanism. This has all kinds of problems. First, it’s not archived publicly. We need artifacts, and research should be conducted as openly as possible. Secondly, it makes it impossible for community members to break off into smaller sized groups because there’s no visibility into what others are doing/finding other than what’s reported by the driver. Lastly, it doesn’t allow for a system of trust or reliability to develop between community members. The problem with the Talking Points Memo setup is that Joshua acts as a bottle-neck between community members.

The Daily Kos site is impressive from a community building standpoint. There’s a moderation system to help protect people from the trolls, a karma system so that trust and reliability can be established, and everything is archived and searchable. These are all requirements for a general system of distributed journalism / activism to work at a large scale.

What’s missing is effective methods for individual community members to find tasks quickly and easily. As it is today, you have to spend a significant amount of time reading through massive amounts of discussion to find actions that match your abilities.

What are some of the things that could be matched?

  1. Geographic location. This is probably the single best attribute. Need me to file some papers at the local government offices? Call my representative? Whatever. In both the Shays handful and the Gannon stories, location alone has been an important aspect in locating potential volunteers.

  2. Skills and knowledge. Do you know anything about Limited Liability Corporation law in the state of Texas? The inner-workings of unions? Investment? Accounting? Internet infrastructure? Finding people with specialized knowledge has been one of the most interesting aspects of the Gannon story.

  3. Access to information discovery services. Will you go to the library and trawl through newspaper archives? Access to Lexus Nexus? Government clearance?

  4. Political and religious affiliations.

Once I can easily subscribe to leads based on my situation, I then need somewhere to go to research the topic (a list of links to weblog entries will do fine) and then somewhere to log my results that I know will be seen by the driver and other community members. I also need the ability to enter more leads that will go through the ability-matching process.

Some Ideas

When I got to this point in the post last night, I started in on a rough draft of a functional specification for a web based system that would facilitate the types of interaction I’ve described thus far. It got out of control very quickly and I decided that I was getting ahead of myself. After reading over it a couple of times, I realized that what I was describing was a combination of a few existing types of applications that I deal with on an almost daily basis:

  1. Common bug tracking software, such as Bugzilla, FogBugz, and Trac.

  2. Weblogs.

  3. Lazyweb. The idea here is that you trackback/pingback a aggregation service with potential issues so that it’s available from a single place (RSS Feed, etc).

  4. Slashdot style commenting and moderation system. This is basically what Daily Kos is.

The general concept would be to allow potential issues to develop the same way they develop today - through weblogs. At some point though, we need to switch over to bug tracking mode where the details of an issue are recorded in such a way that action-to-ability matching can be performed. We need to maintain status, state, and priority on leads, actions, and issues. Each of the leads/actions needs to be recorded with detail on geographical and access requirements. There needs to be discussion space dedicated to each lead/action. Individuals taking up the lead/action need to be able to log their findings somewhere and funnel new leads/actions through the matching space.

Shheewww

Free and open source software development processes have nothing to do with software - we just found them first. I’m revisiting The Cathedral and the Bazarr and finding entirely new ideas when I broaden the concepts to include other types of development. I find this quote by Linus Torvalds to be especially global in its application:

Given enough eyeballs, all bugs are shallow.

Democracy, capitalism, and the press are all machines designed to assist us in reaching our goals. Sometimes I think we forget that they don’t operate on their own. A successful system demands constant maintenance and hacking. If the machine sits for too long, it rusts and becomes a hindrance. Defects and signs of age must be identified, researched, and corrected as quickly as possible. The number one issue facing us today is a serious lack of eyeballs. If we increase our eyeballs, all issues become shallow.