editor's note: as an emerging concept, this page is used both for what is a political wiki and what should be
A political wiki is a wiki which can be used ideally to:
- inform debate on public policy issues
- support deliberation via debate by edit
- enable participatory democracy
Use of wiki is thought to be fundamental to open politics. That is, wikis are the technological base which most closely corresponds to the requirements (see below).
Most political wikis discover early on that they must formalize their online deliberation to structure collaboration, encourage consensus building, accurately reflect dissent and the minority positions. Only then can they expect to become a trusted venue for controversial discussions.
That trust can be lost easily by the errors. A ruleset to minimize those, open politics in force, defines several audit criteria for a political wiki: rootdness, democracy, and so on.
[+] best practices
There are several lists of wiki best practices. The lists most applicable to open politics itself are:
A good political wiki is open and tolerant even to extremes due to a need for neutral (unbiased) referreeing of the inevitable clashes between factions in all political environments. If a faction or clique takes control of the administrative/editorial function, the wiki becomes useless as a consensus or truth seeking environment. The administrator guidelines of a political wiki must be extremely careful not to usurp or anticipate any political judgement of the users.
The users are not using a political wiki to argue with the administrators. They are using it to argue with other users.
[+] examples (from best to marginal)
- openpolitics.ca and its predecessor LivingPlatform.CA
- debateus.org - US site which encourages political dialectic.
- Campaigns Wikia - Jimmy Wales political wiki.
Others have promise but are too lightly used to assess:
- dowire.org wiki
- Green Party of Canada Living Platform lp.greenparty.ca
- Sustainable Nova Scotia, used to generate the ImagineHalifax activist platform
- embodimentwiki.org, used to debate embodiment theories, e.g. cognitive politics
[+] all wikis become political
Simply due to scale, a large public wiki will become a political wiki as more people arrive of different political views and positions. The systemic bias of the current users will tend to be challenged by new users.
Also, all wikis are used politically, by those pushing their own point of view in the honest belief that people who don't accept it on some (or all!) matters, are oppressing or hurting them by not accepting it.
All political wikis change as new users arrive. Usually a dialectic between the neutral point of view held by the existing users, and the New Troll point of view held by new arrivals insistent on the legitimacy of views not held by the insiders (almost always called "trolls") develops.
This may settle into stable tensegrity if some users decide to act as bureaucrats and devote themselves to finding only non-controversial points of agreement between the factions and enforcing them.
Or a sort of ongoing negative campaign may develop, or a sort of representative democracy involving factions, or a consensus democracy relying on arbitration and mediation. There are many ways for a troll-sysop struggle to evolve, but any political wiki will eventually have to face it.
[+] desirable technical features
See open politics service for a longer list that puts these features in context.
- view page - reflects the current version of a document, this is the standard view of a page
- edit pages to create a new current version.
- page history enables viewing and comparing of all versions.
Any user page, talk page, project page, follows the same protocols, though the users of these types of pages differs radically.
[+] consistent commands
A list of all command verbs is required both for translation and consistent use within the text. This is a web best practice that is just as applicable to wikis. It is even more applicable to voice services such as Voxable.
If it is important for people working in multiple languages to participate, then, the user interface must demonstrably not favour any one mother tongue. Another reason to maintain the list of all command verbs is to ensure that the translations are correct and suggest the correct conceptual metaphors, so that users of both languages proceed on an equal power relationship basis.
This is an extraordinarily difficult guarantee to achieve, and will probably cause most multilingual wikis to fail in a political context, or be restricted to fairly primitive means of decision making. Wikipedia does a reasonable job of this with consensus democracy techniques, but only its English version really applies these with any rigour.
[+] activity, identity, attribution, authentication, permission
People who are less than expert in identity, attribution, authentication and permission security simply cannot create an access control structure of any reliability. This is an expert task.
Most wiki software is free software and not only does not reflect the modern understanding of these problems, but, proliferating versions of these ensure that they never will: a version may be robust in one release, fail in the next, and few (except those finding the exploits) will notice.
Accordingly some developers including Steven Ericsson-Zenith believe that there is no future for free software for sensitive content management, and have quite strong arguments. By constrast the mass peer review approach has produced a more reliable OS than Microsoft's by most measures, though it takes time to do so, and this technique may yet yield a wiki that is up to the many challenges:
[+] complete activity and changelogs
[+] support for anonymity
There is controversy and confusion on the question of authentication, all forms of identification but especially outing by others, on the uses of pseudonyms, and anonymous edits.
Politics itself operates on a basis of incomplete and partially-leaked sources, which is why people working as journalists must validate sources and double-check assumptions.
The dominant view among political bloggers and wiki users that information provided by anonymous sources must never be traceable back to any person, ever. That opinion is ultimately separable from the body sharing it. All persons have the right to assert this separation in any liberal democracy today. Only South Korea has attempted to pass laws to prevent this separation and that only for domestic users.
The reasoning is not hard to trace. Under a repressive regime unconcerned with human rights, even innocuous information could be dangerous to the individual. Countries with strong rights protections tend to protect privacy as one of the most important rights of all, due in part to the need to distrust all forms of authority by default.
A very few argue that a liberal democracy, and a justice system must require an amount of personal identification and accountability on opinion. They may analogize the situation of discoursing on opinion to that of engaging in contracts and may propose a collective social contract to be enforced by some arbitrary and universally trusted authority. By this logic, protection of anonymity is only a right when it is in the public interest to grant it, in the opinion of (only) the dominant local authority.
As a simple example, a person posting a pro-homosexual position could be arrested and tortured on entering a country that takes an extremely homophobic attitude. It's happened. And if someone doesn't know their identity has been compromised, they may well enter such a country willingly. The service in such a case might even liable for the leak of the data, in some countries, restricting its activities in those countries or ability to raise funds in those countries.
One of the fatal tikiwiki flaws is that it takes a lax attitude to this protection. It is quite difficult to suppress IP numbers wherever they appear. However, it's mandatory to do so.
A more complex example involves an opinion that is considered subversive in one place but innocuous in another. For instance, a person expressing the opinion that a political party should not give posts to its major donors or lenders, or be otherwise "bought" by them or "sell out" to them, may be subject to libel laws in one jurisdiction and assassination in another. However, if the public interest is weighed only in the jurisdiction that hears a civil lawsuit, the interests of the public in other places where these events occur will not be well considered.
Regardless of technical support, social support remains necessary:
Anyone revealing information not available to all users must reveal to all users that they have done so.
When outing other users this is particularly important.
[+] editorial attribution
Editorial authority can be vested in accounts (this is not the same thing as bodily authority, and not the same as sources, see biometrics below). The use of a login to approve content or verify its status is common in all large public wikis, though there is usually no attempt to verify an individual's identity.
As identity theft is common, a delaying and staging procedure is recommended taking at least a week before something can be considered to be accurate or published:
[+] hardwired notices
The flipside of this requirement is that there must be no confusion about the authority that is behind such evidence: there is none, it is speculative until someone qualifies and states the nature of the material. For instance, a draft press release may be just that, and it may contain unapproved quotes or unverified facts. The notices that tell end users to quote or trust this information must be clear on the status of the information. For instance any page touched only by anonymous trolls should contain a notice such as the following at top:
- This page has been edited only by anonymous trolls of unknown motives and trustworthiness. Do not quote from it. Do not assume it correct. Even if it appears correct it may contain facts or quotes that are unverified or inaccurate. No regular editor at openpolitics.ca itself has verified or paid any attention to this yet.
As others who accept the assertions add detail, it should escalate in reliability:
- No senior editor has verified this page. While regular editors have edited it they are not vouching for its whole content as accurate or from reliable sources.
- One senior editor has edited this page, it may be considered useful enough to ask questions about. This is not the same as quoting from it.
- The chief editor has verified that all of the quotes and facts on this page are essentially correct. If you object, you have one week to do so, else it will be considered to published and will be made available for any and all uses.
- This page has been published. This means that all liability for its content has been accepted by openpolitics.ca itself, and all sources consulted to verify it are protected to the same degree as any journalist's source.
This is also the correct procedure for Wikipedia and other large public wikis, though they are too stupid to realize it and have adopted counter-productive measures that actually make it far easier to spread rumours and far more difficult to track those spreading them.
[+] rendezvous, migration, faction, advocate and collective identity support
Individuals in democratic politics have a right to political privacy, e.g. secret ballot, and freedom of association, meaning, it is not ever acceptable in democratic politics to attempt to prevent communication between persons - see communication ban. People should be able to rendezvous and migrate services without revealing their body identity to anyone. This is called a thidwick in wiki circles.
[+] factions, parties and other groups
Also, as politics itself is a process whereby groups of people assert themselves as groups and curtail individual excesses, e.g. by defining them as crime, it is also unacceptable to let one or more visible members of a group take more than the amount of revenge or feedback that they have volunteered for. Groups must be able to stand together for purposes of sharing a view, e.g. in a political party. It must also be possible to entreat or sue the group or entity. The reverse is not acceptable however: groups or parties must not have the right to single out and sue or harass individuals, for various reasons. This has been established in Canadian courts to a limited degree by lawyer Julian Porter.
Finally, persons have a right to let issues be heard out in private amongst themselves before committing themselves to an outsider's view. It is entirely reasonable for any entity other than a political party to have private space and time to decide what is their basis for unity, for instance. And, if there is a dispute on the accuracy of some article about Karl Marx, it is reasonable to require that someone who is at least familiar with that doctrine review it. Groups must also have the right to advice such as legal advice, meaning they can have advocates speak for them with respect to a given matter and prevent other communication from having status. Though this may have severe consequences within the group.
The typical word for features that let a group act and speak as a group is a faction. At present no wiki support this directly, but some have used interim means such as shared accounts. This is just another role account which is used to speak for the faction itself and which can only be used by rules set by the faction itself. While enforcing these rules is beyond the scope of the wiki, or administrator guidelines, it is possible to break down what a faction or party does into organization protocols and require that some of these be published so that outsiders can verify that the group has in fact "acted as a group" with respect to any given matter.
For instance, a group that wants to be represented by its lawyer, only, would have an official contact with that lawyer, and means to verify that the lawyer was in fact acting for the group.
Any service that claims to respect journalistic standards will have to have a means to suppress or deny identity claims made on the basis of specious information or suspicion, e.g. known associations, IP addresses, shared logins and open secrets.
At the very least, participation in the service must be denied those who speculate on identities of persons who are not seeking political power or directly engaged in the seeking of such power, e.g. as political party officers.
A political party officer or candidate or government employee has no such right to privacy, however, as they influence and control the legal structures and systems. A political wiki is a means to balance that, and so treats critics of such people asymettrically.
The use of an ordinary login to attempt to verify an individual's identity is subject to some extreme disclaimers. It is routine for instance for spam to be sent from zombie accounts via zombie PCs, hijacked by relatively simple means. Use of proxy servers, password swaps and identity theft is also common.
If nothing else, laptops and BlackBerry and other PDA devices can be stolen and "cracked". This may become for instance increasingly common during elections, if this is indeed the Troll Age, as Craig Hubley claims.
Given all that, requirements like "users need to be logged in for the purpose of verifying one's identity" are fatuous. No login process can be made secure and reliable. At best a balance of probabilities degree of verity can be assumed.
Even a PGP signature is inadequate for some, e.g. military or emergency, types of notifying, that will result in immediate and irreversible action. For such applications, at least a three party quorum key is required so that any trust granted a particular ID can be very quickly revoked.
If the above fails for some reason, there is only one alternative.
To operate beyond a reasonable doubt as would be appropriate for criminal law would require biometric authentication such as available in systems like memeio.
Some laptops from IBM now have biometric identification built in. These or some other standard single device must be mandatory for any organization relying on biometrics, so that all use of biometrics in an organization has single point of failure.
[+] scalable permissions feature
The essential functions of a typical permissions structure include:
- defining categories of objects.
- defining groups of users.
- applying permissions to groups
- applying permissions to categories
View/edit/history permissions must always be separable.
Permissions must apply across all features.
Permissions shoudl not be applied to individual objects.
Projects requiring logins may also support a permissions structure to prevent viewing or editing. In most wikis this is very minimal with only the ability to lock pages. Politically, it is often argued that permissions are a form of prior restraint and must be used only in rare instances and temporarily or to preserve basic operations and instructions for regular review (between which, they are on locked pages). For instance, a page saying officially who the leader of a political party is should be locked until such time as the leadership race begins to avoid speculative information on rivals being worked in without some kind of central editorial.
The problem with permissions structures is that one must be not only consistent with the governance model, but wholly determined by it, in any democratic organization:
That is, if the organization's constitution or corporate bylaws establishes that only a certain officer sees certain material, it must be possible to ensure via some means such as a quorum key, that all members do in fact trust or believe. That trust existing, it must then be easy to verify that these measures were employed:
All use of any feature that is restricted by a permission structure must be by one of two means:
- role accounts that provide that access, e.g. "root", "final legal vetting before publication" and whose use is fully and reliably logged
- body named accounts that identify a person directly, e.g. User:Craig_Hubley.
An ordinary troll name or pseudonym is not effective, or adequate, except in the context of faction support where the faction (not the service) has agreed to be bound to that account and take full responsibility for all its actions.
Creating a fully governed permission structure is an expert task and requires multiple governance experts to be involved. A typical political party might have to devote several months to the task, and put at least three or four very trusted people on it. Its political party constitution would have to be revised to make very specific mention of the roles involved and would at least have to list organization protocols.
Any lesser effort has proven to result in a tension between technical and political power structures, and a sort of troll-sysop struggle between them until the technologists control the technology, and the politicals control the money, and never the twain shall meet. See Living Platform in Practice for evidence.
instructional capital to make social capital, it takes some social capital to start with, like sand in an oyster to make a pearl:
Creating social institutions worthy of trust requires the involvement of those with skill, and this can only be obtained by gradual learning. If you want to create a technocracy, this can be achieved using externally certified experts, who operate a command heirachy of their own design.
For the purposes of democracy, it is better to create learning hierarchy, which identifies learning pathways, levels of achievement, and explains training (including self training) that is available. Rather than trying to define "trust" or authority into existence, it is better to observe and describe what evidence/source/authority is actually being cited in the context of real decisions. And, as noted above, rather than trying to prevent errors or manipulations by prior restraint and permission, it is better to operate by disclaimer and solve only problems that actually materialize.
reflexive governance structure is one which
- A. does not rely on outside authority for legitimacy
- B. is internally consistent, i.e. if it claims to be a participatory democracy then no aspects of the governance model are excluded. Governors or administrators must eat ones own dog food, rather than use the administrative fiat.
The most successful political wikis have been very reflexively governed by their own users.