The Internet, Policy & Politics Conferences

Oxford Internet Institute, University of Oxford

Heather Ford: The socio-technical shaping of collective deliberation in Wikipedia: The case of the 2011 Egyptian Revolution article

Heather Ford, Oxford Internet Institute


Wikipedia policy states that consensus is the primary way decisions are made on Wikipedia and that consensus ‘on Wikipedia does not mean unanimity (which, although an ideal result, is not always achievable); nor is it the result of a vote’. Rather, ‘(d)ecision-making involves an effort to incorporate all editors' legitimate concerns, while respecting Wikipedia's norms.’ (Wikipedia authors, 2014)

Wikipedia’s consensus building is said to be a natural outcome of the editing process: someone makes an edit to a page and someone else can then choose to keep it as is or to change it. This image of consensus building is at the heart of the majority of scholarly accounts of Wikipedia governance. As scholars of decision-making will note, the extent to which decisions are actually consensual are based on the details and context of the decision-making process.

In the highly mediated context of Wikipedia, it is useful to try to understand decision-making in practice from perspectives of materiality. In this paper, I argue that decisions in Wikipedia are made under the material conditions of particular socio-technical systems; they are influenced by the affordances of the software and of the material dimensions of the discussion space that outline what can be discussed or not discussed. 

In practice, a range of decision-making techniques that have particular material dimensions is used to propel the project forward. When editors continue to disagree with one another when editing or when they want to propose a change that they believe will have significant effects, they discuss this on the talk page. The editor starting the discussion will set the dimensions of the discussion by asking for a poll or by framing the issue as a question, for example. If they cannot reach consensus on the talk page, then editors can ask for wider deliberation through third opinions and requests for comment on other spaces in the encyclopedia that have their own system for dealing with consensus. More extreme processes that will take authoritative (rather than consensual) steps to end the dispute take place in judicial-like processes including administrator intervention, formal mediation and arbitration where affected parties are asked to make statements in their defense.

Beyond this, the Wikimedia Foundation regularly makes authoritative decisions about issues that they believe have a legal impact on the organization as the institutional host of the project that are outside the purview of editor consensus. Jimmy Wales, as founder of the project, also has a ‘benevolent dictatorship’ role (used less now than in the earlier years) where he is able to make unilateral decisions to appoint administrators (after a general vote) and ban users. In these cases, there are few (if any) dedicated spaces to discuss these issues and so users must “take over” certain spaces in order to voice their disagreement.

In this paper, I develop a taxonomy of decision-making techniques by following edits and discussions in the Egyptian Revolution article according to three of the major disagreements that occurred, including: naming the article, tracking death counts and selecting and captioning images.

Editors used numerous methods of decision-making including polling and different types of discussions, but they also imposed bans on what they saw as unruly editors and they protected the page from anonymous editors at times when the ‘war came to the wiki’ as one editor complained.

The goal of the paper is to complicate the often taken-for-granted notion of Wikipedia’s consensus building in order to answer the following questions: What kinds of decision-making techniques are used to resolve disputes and how are they materially enacted in Wikipedia’s socio-technical system? Wikipedia’s governance system has been widely lauded as a model for other public information goods but more research needs to be done to understand how governance and consensus play out in crowdsourced environments. 

Heather Ford