Consensus is part of a range of policies on how editors work with others. Consensus is not what everyone agrees to, nor is it the preference of the majority. Consensus results in the best solution that the group can achieve at the time. Remember, the root of "consensus" is "consent". This means that even if parties disagree, there is still overall consent to move forward in order to settle the issue. This requires co-operation among editors with different interests and opinions.
As a decision-making process, consensus decision-making aims to be:
- Inclusive and participatory: As many editors as possible should be involved in the consensus decision-making process.
- Cooperative: Editors in an effective consensus process should strive to reach the best possible decision for the wiki and all of wikians, rather than opt to pursue a majority opinion, potentially to the detriment of a minority.
- Egalitarian: All editors in the consensus decision-making process should be afforded equal input into the process. All users' opinions have an equal weight, and have the opportunity to present and amend proposals.
- Solution-oriented: An effective consensus decision-making process strives to emphasise common agreement over differences and reach effective decisions using compromise and other techniques to avoid or resolve mutually-exclusive positions within the group.
- Most logical: This happens when a solution appears to be impossible to execute because of the lack of support and co-operation.
Consensus discussions should always be attempts to convince others, using reasons. When a discussion breaks down to a mere polarised shouting match, there is no possibility of consensus, and the wiki will suffer. That said, consensus is not simple agreement; a handful of editors agreeing on something does not constitute a consensus, except in the thinnest sense. Consensus is a broader process where specific points are considered in terms of the wiki as a whole, in the hope that editors will negotiate a reasonable balance between competing views.
In general, the steps involved in the process of decision-making are as follows:
- Discussion: The issue is discussed with the goal of identifying opinions and information on the topic at hand. The general direction and potential proposals for action are often identified during the discussion.
- Formation of a proposal: Based on the discussion a formal decision proposal on the issue is presented.
- Call for consensus: The original proposer calls for consensus. Each editor must actively state their agreement/disagreement with the proposal, to avoid the group interpreting silence or inaction as agreement.
- Identification and addressing of concerns: If consensus is not achieved, the dissenters present their concerns on the proposal, potentially starting another round of discussion to address or clarify the concern.
- Modification of the proposal: The proposal is amended, re-phrased or ridered in an attempt to address the concerns. The process then returns to the call for consensus and the cycle is repeated until a satisfactory decision is made.
Although the consensus decision-making process should ideally identify and address concerns and reservations early, proposals do not always garner full consensus from the wiki.
When a call for consensus on a motion is made, a dissenting editor will have these options:
- Neutral/Comment: Editors who are willing to let a motion pass but desire to register their concerns with the group may choose to "comment". If there are significant reservations about a motion, the original proposer may choose to modify or re-word the proposal.
- Oppose: Editors who are unwilling to let a motion pass without modifications to proposal may choose to "oppose". The proposal will have to be modified or re-worded, or editors may choose to submit an alternative proposal, or a compromise (a middle way between two extremes).
A healthy consensus decision-making process encourages editors with differing opinions to post their concerns as early as possible, maximising the chance of accommodating the views of all minorities. Since unanimity may be difficult to achieve, especially in large communities, or unanimity may be the result of coercion, fear, undue persuasive power or eloquence, inability to comprehend alternatives, or plain impatience with the process of debate, an alternative benchmark of consensus is used: rough consensus.
Rough consensus is a process with no specific rule for "how much is enough." Rather, the question of consensus is left to the judgement of a non-partisan administrator or bureaucrat who has not voted on the issue.
The means to establish rough consensus was described by the IETF (1998) as follows:
|“||Working groups make decisions through a "rough consensus" process. IETF consensus does not require that all participants agree although this is, of course, preferred. In general, the dominant view of the working group shall prevail. (However, it must be noted that "dominance" is not to be determined on the basis of volume or persistence, but rather a more general sense of agreement). Consensus can be determined by a show of hands, humming, or any other means on which the WG agrees (by rough consensus, of course). Note that 51% of the working group does not qualify as "rough consensus" and 99% is better than rough. It is up to the Chair to determine if rough consensus has been reached.||”|
— IETF Working Group Guidelines and Procedures Wikipedia:Rough consensus
Note: While there are no magic numbers for "how much is enough", it is generally accepted that rough consensus has been achieved if about 70% agree on the issue.
In any discussion on this wiki, there are two possible outcomes: consensus, for some course of action, and no consensus. When there is no consensus on the proposal, the current status quo prevails. In the event where no consensus is achieved, it becomes the responsibility of the dissenting editor(s) to modify the original proposal if they wish to. However, it is also possible for the proponents to come up with an alternative proposal which addresses the concerns raised by the dissenters.
It is important to note that a few vocal dissenters do not create "no consensus".
Silence and consensus
In some cases, consensus can be presumed to exist until voiced disagreement becomes evident. One can find out whether an edit has consensus when it sticks, is built upon by others, and most importantly when it is used or referred to by others. However, in discussions regarding policy change, silence may not necessarily imply consent.
Most of the time, editors will find that it is fine to assume consensus, even if just for now, as it is more important to keep editing and co-operating smoothly in good faith as much as possible.
When real people are affected by a decision, such as blocking users, positive confirmation is preferred. Even in these cases, however, dissent might show up later, and it is then no longer appropriate to assume consensus.
Consensus may or may not be reached, but either way, the discussion needs to be closed in the end. The discussion may only be closed if one or more of the following has occurred:
- The topic has unanimous or near unanimous support or opposes and discussion has ran for at least a week or two.
- Legitimate discussions must remain open for a minimum of one week.
- The topic's arguments have been refuted and there is general consensus on which is the better and more accepted option.
- The topic has been open for a long period of time and the discussion has reached a standstill with no sign of consensus forming. The topic should be closed, on a basis of no consensus.
- The topic is listed on Forum:Yew Grove/Previously rejected proposals (in a Yew Grove thread's case) and no new arguments have been brought up, or the author is unaware that it is listed.
- The topic does not affect most or the whole community and is irrelevant and/or unnecessary.
- The topic is meant to be in another area of the wiki, such as the RuneScape:Requests for deletion.
Only administrators may close discussions, although all users are welcome to request closure, if they believe one or more of the above has occurred. Other users may support, oppose or comment on the closure, and an administrator will eventually decide on what action to take, using the above guidelines.
If you disagree or feel uncomfortable with a proposal or its alleged repercussions, the burden is on you to say so.
It is impractical to wait forever for affirmation: in the meantime, it is best to assume that silence implies consensus. Users can continue to hold that assumption (hopefully safely) until someone comes along and voices their concerns. The more visible the statement, and the longer it stands unchallenged, the stronger the implication of consensus is.
Silence is the weakest form of consensus
The wiki is huge and our editors' time is limited. At any given time, there are many open discussions on many different topics across the wiki. It is always encouraged that editors be bold. It is highly likely that editors will eventually find themselves affected by the outcome of some decision that they did not know about, or did not have the chance to join. Where a decision is based mostly on silence, it is especially important to remember that consensus can change (see Consensus is not permanent).
Canvassing is sending messages to multiple wikians with the intent to inform them about a community discussion. Under certain conditions, it is acceptable to notify other editors of on-going discussions, but messages that are written to influence the outcome rather than to improve the quality of a discussion compromise the consensus building process and may be considered disruptive.
Campaigning is an attempt to sway the person reading the message, through the use of non-neutral tone, wording, or intent. While this may be appropriate as part of an individual discussion, it is inappropriate to canvass with such messages.
Votestacking is an attempt to sway consensus by selectively notifying editors who have or are thought to have a pre-determined point of view or opinion (which may be determined, among other ways, from a userpage notice, such as a userbox, or from user categorisation), and thus encouraging them to participate in the discussion.
In the case of a re-consideration of a previous debate (such as a "no consensus" result), it is similarly frowned-upon by many editors to send mass talk messages to those who expressed only a particular viewpoint on the previous debate.
Posting a friendly notice on users' talk pages in order to inform editors on all "sides" of a debate (e.g. everyone who participated in a previous debate on a given subject) may be appropriate under certain circumstances on a case-by-case basis.
Because it is less transparent than on-wiki notifications, the use of email or other off-wiki communication (in-game messages) to notify editors is discouraged unless there is a significant reason for not using talk page notifications. Depending on the specific circumstances, sending a notification to a group of editors by email may be looked at more negatively than sending the same message to the same group of people on their talk pages.
The term "forum shopping", or "asking the other parent", refers to repeatedly asking for additional outside opinions until a desirable opinion is obtained. For instance, if the user is blocked, they can ask for an outside review of said block; if this review concludes that the block was proper, it is generally inappropriate to repeatedly continue to ask for yet another outside review.
This also includes bringing up the same issue at the same forum multiple times, or asking multiple administrators to perform an administrative action after an administrator already decided against it.
Consensus is not...
A majority vote
Consensus is not a majority vote. Voting has (at least) one inherent flaw in use in wikis, and that is that votes are semi-permanent, and community discussions are not.
A community discussion shifts with the community members. When a community view changes, its approach will merge seamlessly into the new line of thinking. This is because a community view is intangible. Votes, on the other hand, are nailed to the wall – and, worse still, the best way to really be rid of a vote is to have another vote against it.
Votes are especially unfair to those who come after. Someone walking in on a situation in which the dice have already been cast do not even get the chance to vote.
Every opinion counts. Consensus accounts for dissent and addresses it, although it does not always accommodate it. An option preferred by 51% of people is generally not enough for consensus. An option that is narrowly preferred is almost never consensus.
A vote may help to organise discussion around specific proposals, but this can sometimes breed conflict and division. One problem with a yes-or-no vote on a proposal is that there may be a consensus for a middle option. Even a "middle ground" option can be insufficient, as forcing people to choose between options may prevent new ideas from coming forward that would gain more support. Another problem with voting is that it might prevent a real discussion, as voters do not have to justify their position. This prevents people from evaluating the underlying reasons for a vote, and criticising weak or inaccurate reasoning for a vote. It also prevents people from coming up with alternative ways to satisfy the voter's concern, with a less divisive course of action.
The best way to determine consensus is to actually read and understand each person's arguments, even if they are divided on the surface. A consensus can be found by looking for common ground and synthesising the best solution that the group can achieve at that time.
Consensus is not the same as unanimity. Every discussion should involve a good faith effort to hear and understand each other. But after editors have had a chance to state their viewpoint, it may become necessary to ignore someone or afford them less weight in order to move forward with what the group feels is best. Sometimes a rough consensus is enough to move forward (see Rough consensus.)
Insisting on unanimity can allow a minority opinion to filibuster the process. If someone knows that the group cannot move forward without their consent, they may harden their position in order to get their way. This is considered unacceptable on this wiki as a form of gaming the system. There is even a three revert rule to limit efforts to stone wall the editing process.
Editors should make a good faith effort to reach a consensus. That means that the dissenting party has to state how the current proposal fails to meet the interests of the wider group, rather than merely stating they will not accept it. But after a good faith discussion, sometimes the dissenting party must consent to move forward even if they disagree with the specific course of action.
All or nothing
If the group can identify areas of agreement, they should move forward where the group shares the same view. A complicated dispute might involve several issues, and some issues may be more controversial than others. But a disagreement on one issue should not prevent consensus on another issue. It is not helpful to expect complete and total agreement on every aspect of the dispute. Work with the issues where there is common ground, and revisit the lingering issues later if necessary.
Consensus is not immutable. Past decisions are open to challenge and are not binding, and one must realise that such changes are often reasonable. Thus, "according to consensus" and "violates consensus" are not valid rationales for making or reverting an edit, or for accepting or rejecting other forms of proposal or action.
The wiki remains flexible because new editors may bring fresh ideas, growing may evolve new needs, editors may change their minds over time when new things come up, and we may find a better way to do things.
A representative group (i.e. administrators and bureaucrats) might make a decision on behalf of the community as a whole. More often, editors document changes to existing procedures at some arbitrary time after the fact. But in all these cases, nothing is permanently fixed. The world changes, and the wiki must change with it. It is reasonable and indeed often desirable to make further changes to things at a later date, even if the last change was years ago.
A walled garden
Even where there is a consensus among a group of editors, their preferred outcome is not always acceptable on the wiki. For example, a decision made on an article's talk page by a small group of editors does not override community discussions and consensus in the Yew Grove.
Editors can easily create the appearance of a changing consensus by "forum shopping" (see Forum shopping): asking again and hoping that a different and more sympathetic group of editors discusses the issue. This is a poor example of changing consensus, and is anti-thetical to the way this wiki works. The wiki does not base its decisions on the number of people who show up and vote; we work on a system of good reasons.
At the same time it is normal to invite more people into the discussion, in order to obtain new insights and arguments. However the invitations must be phrased in a neutral way and addressed to a reasonably neutral group of people (e.g. sent to all active editors or posted at the SiteNotice).
It may be tempting to solicit opinions from wikians or administrators who agree with your viewpoint in order to get your way. It may also be tempting to ignore the consensus found at one forum, and solicit a new discussion at another forum. This violates our behavioural policies and guidelines. Sometimes it is appropriate to try a different dispute resolution process after one has failed. But there is a difference between reasonable dispute resolution and gaming the system, and it is important that wikians understand that difference.
While everyone on the wiki has the right to be heard, this does not mean that discussions remain open indefinitely until we hear from them. Nor does it mean that a consensus should be overridden by an appeal to "wikians out there" who silently disagree. In essence, silence implies consent. If the current discussion does not represent real opinion, either prove it by referring to an existing discussion, or suggest starting a new discussion with a wider audience.