Spending another flight on Socracy

As I get a little bit more into the details, the requirements on the general structure of the ‘thinktank engine’ of Socracy are getting more complex. Flight after flight, I have been able to work a bit more of the list of topics, exploiting mainly the political section of newspapers and making sure that all major topics discussed could fit within the defined nomenclature. A while ago I also uploaded all political programs available on the web, and enriched the structure to ensure that proposals could be inserted. The next work will be to attempt to insert the list compiled by the Thomas More thinktank (institut-thomas-more.org), which contains no less than 490 proposals made by Nicolas Sarkozy during the last presidential election.

Indexed by the nomenclature of themes and keywords, the content itself will be of two types: facts and proposals.
The fact documents will need to contain factual & verifiable data concerning the theme. e.g. statistics, existing laws, historic events, etc. Editorial rules should similar to those of Wikipedia, with an obligation to mention the source and if possible a link to that source if available on the web. The fact documents contains HTML text and provide a storage place for attachments. They can be edited by anyone (e.g. using recent press articles) and therefore use a regular Wiki engine, which ideally should provide the following features:
 flagging a fact document as not passing quality requirements (lack of sources, inaccurate source or interpretation of the source)
 Subscribing to all updates to this facts document
 add this facts document to your favourites
 view last updates (part of revision management)
The proposal documents will be many for each theme entry, as obviously it would be unrealistic to expect all stakeholders to agree on a common document. Socracy simply aims at facilitating the confrontation and convergence of political opinions by ensuring easy access to each other’s proposals, and allowing opinion leaders to comment and challenge each other’s proposals.
This constraint to have multiple proposal documents for each theme is an important difference with Wikipedia. The ordering of these proposals will be important to ensure that ‘dominant’ opinions, reflecting the widest possible consensus, get presented first. How to assess which opinions are dominant is indeed a problem. The idea is to allow edition of a proposal document only by members who first indicate that they “agree overall” on the proposal. This discourages opponents to edit a proposal in an unconstructive manner, as they would also, by the same token, promote it. The default ordering of proposals would therefore be by number of members who indicated they “agree overall”. If a member does not agree with any proposal and has another one, he can and should create a new proposal document.
For each proposal, the Socracy engine should provide the following features:
 Display of current number of contributors
 Display list of supporting named contributors and parties.
 Display of geographical scope (e.g. a city, a country, international)
 Display of proposal HTML text
 Display of proposal attachments
 add this proposal to your favourites (in which case it appears first for you)
 Flag this proposal as having inappropriate content
 Contribute to the proposal (you will add to the number of contributors for this proposal)
 Subscribe to contributions for this proposal
 If you are the owner of the proposal (original contributor):

  • lock/unlock the program (if locked you will not allow people to contribute). If the program was contributed by a virtual party, then all members of this virtual party have access to revision
  • revision management(Rollback notifies contributor). management even for locked programs.
  • Assign proposal to a virtual party

After gathering information on popular Wiki engines hoping to find one that would get close to the requirements, using a combination of Google and Wikipedia, I became convinced that XWiki would be a good candidate, being written in Java and offering extensive features. Still of course no existing Wiki engine was tailored to the needs of Socracy, and modifications would be required: I had to contact the team of contributors to assess if such changes were possible
Soon I discovered that XWiki had been initiated by Ludovic Dubosc, who had exactly the same curriculum as mine, and was leading the company behind XWiki with most of the programming team in Paris ! This would make things even more simple.

On the topic of Politics & the Web, Ludovic happens to be an expert, as he contributed to the back-office of one of the candidates during the last presidential election and also had thought of leveraging Wiki tools to better expose programs. The main obstacle, according to him, was to avoid spam. It seems that the favourite sport of political parties during campaigns is to spam the sites affiliated to opponents. A logical conclusion is that this rules out the possibility of anonymous membership on Socracy, which is a big problem as political opinions, especially in France, are often perceived as something extremely personal, and I expect that many of the experts I would like to contribute to Socracy would not feel comfortable to be exposed. You can be an expert but not a militant !
After a bit of thinking on this issue, I believe we can allow 2 types of members on Socracy:

  • Identified members. For such members Socracy knows verifiable identification details (email, phone number). These members are expected to be those most actively involved in politics, already exposed and who simply extend this exposure to Socracy. Of course Socracy never exposes such private details and these members appear only under a pseudo of their choice. However they will have the special mark and privileges of an identified member.
  • Referral members. These members do not leave any personal detail to Socracy when they subscribe, but they need to be approved by an identified member. For instance most think-tanks are animated by a team of people actively involved in politics and who probably will agree to be identified members, but other contributors to the think-tank may just ask them to ‘sponsor’ they membership on Socracy. In case a referral member shows rogue behaviour, it remains possible to contact the referral member via his sponsor, so it is expected that referral members will not spam.

Members can get organized into virtual ‘parties’. A virtual party is simply a group of members. A virtual party is created by a member, the lead, who can accept and revoke party memberships (he cannot arbitrarily put members in his party, he can only invite them and accept membership applications).
Proposals can be initiated by a member under his own name, and he can assign the proposal to any virtual party that he leads. Each virtual party models a think-tank, and there is a possibility to limit proposal edition only to members of a virtual party.

Last but not least, it should be possible to access Socracy in ‘Party view’ only, a mode where ONLY proposals of a given virtual party are visible. Why ? Obviously Socracy will fulfil its mission of facilitating confrontation of proposals only if opinion leaders contribute their proposals. Most of them also have their own web site. Contributing to several web sites can be cumbersome. The idea is to propose Socracy as the engine behind the party/thinktank “proposal” section of their web own site, which will be accepted only, of course, if it displays only content relevant to the party.
I believe the timing is correct for proposing such an engine in “ASP” mode (but at no cost of course), because most parties have started to feel that they need to open-up a bit more to proposals of their members, to adopt an approach which would be less strictly ‘top down’. The attempt to “policy making 2.0” so far have proven to be a bit too ad-hoc, with an obvious lack of efficient tools. Socracy should aim to address this pain point and become the tool for policy making ‘à la’ 2.0.
At local level, more modestly, I do see signs that demand for such a tool is building. Parents in the school of my kids are organized in 2 “lists”, and are wondering about launching a blog to gather concerns on school life and expose their proposals, while the mayor would like to also contribute his proposals. It seems that local politicians have become concerned by the fact that local associations too often represent only their leader, a vocal person not necessarily representative of the general opinion. As only opinions of such associations get exposed, they run the risk of being misguided, paying the price at the next poll. Socracy can be an efficient tool to organize a better feedback process. Although proposals can be contributed by a virtual party, the weight of the proposal, when it is contributed, is only 1, not the total number of members in the virtual party. Supporting a proposal requires an action, and ensures that only virtual party members truly interested on the proposal count as supporters. In order to further ensure that supporters do correspond to active members (avoiding temptation by parties to create fake referral members), a supporter of a proposal will be counted only if he is ‘active’ (to be defined, probably last login to Socracy in the X preceding month).
Socracy, therefore, could become an embedded tool not only within the web sites of think-tanks or political parties, but also local administrations. The referral membership mechanism will enable a party to manage his members in a totally autonomous way, Socracy being even unaware of the identities of members. There may be a need to avoid duplicate referral memberships across different sponsors, but this is likely to be too marginal to really affect the engine.
To move all this forward will require more long flights and a lot of programming…