Universal Code of Conduct/Initial 2020 Consultations

From Meta, a Wikimedia project coordination wiki
Jump to navigation Jump to search
Other languages:
English • ‎dansk • ‎español • ‎русский • ‎हिन्दी • ‎日本語

Introduction[edit]

In June 2019, the Trust and Safety team conducted a study on the existing behavioural policies in the Wikimedia movement. The results indicated that over 50% of the Wikimedia communities have underdeveloped policies[1], or no policies at all, for conduct guidance and local enforcement.

To better understand the working atmosphere of communities across our movement, T&S launched a three-month-long local language consultation process in a number of very different communities. The consultation was done with the help of 12 community members who engaged with 19 communities in their local languages in order to understand the status of behavioural policies and discuss whether they would benefit from having a universal code of conduct (UCoC) for the movement. The main purpose was to obtain qualitative insight on communities’ needs and concerns in regard to conduct policies.[2] 

T&S considered a variety of factors in choosing the specific languages for consultation. We wanted to hear opinions of Wikimedians from different parts of the world. We wanted especially to bring voices into the discussion that usually are not heard in central, English-language-dominant channels, like Meta or mails in Wikimedia-l (in the past, smaller and non-English speaking communities have expressed their hesitation in participating in global discussions for fear of not being heard, so we wanted to avoid that ourselves). And we wanted to hear perspectives from diverse communities: smaller and larger ones, newer and more established communities as well as from communities with different levels of development regarding behavioural policies. In the end some of our choices were also based on practical considerations like the languages spoken by qualified applicants for the facilitation.

Demographics covered[edit]

Regions covered in UCoC consultation

Engagement[edit]

UCoC comprehensive summary - Engagement with different communities.png

The feedback from the communities was diverse with varying response rates. A few of the communities actively participated in the consultation process while others had to be persistently pursued. Some were reluctant in the beginning but gradually opened up while others remained closed throughout the process. Furthermore, COVID-19 related concerns and associated real life challenges impacted the ability of community members to engage. In some communities this impact was heavier than in others. Whether the communities were enthusiastic about the discussion or not, they, for the most part, voiced appreciation for the Foundation making an attempt to engage with them in their local languages. They insisted that should a code of conduct be rolled out, it must be translated into multiple languages.

Medium of engagement[edit]

UCoC comprehensive summary - Response from various communication channels.png

The medium of engagement was determined by the needs and practices of communities and individuals.  Different language communities have different preferences for discussion methods and communication channels. Some communities like to conduct important discussions on Meta whereas a majority of them prefer to have discussions on their village pumps with some side discussions on social media channels. Since a discussion about harassment cases and behavioural issues is a sensitive topic for many community members they voiced their choice of having those discussions either through direct private discussions or sending their responses through anonymous surveys. The facilitators respected the voiced needs and utilized the preferred conversation spaces and styles of their communities. They used a variety of methods to engage with the communities and adapted to new channels as and when requested by community members.

Results[edit]

UCoC comprehensive summary - Distribution by gender.png
UCoC comprehensive summary - Distribution by rights.png

The feedback from the communities has been duly recorded and summaries have been written by the language facilitators. Those summaries are independently presented under specific language headers, with some languages combined into one report.

Insights from the communities:[edit]

  1. Arabic
  2. Bengali
  3. Chinese
  4. Hindi
  5. Japanese
  6. Kannada +Tulu
  7. Malayalam
  8. Odia
  9. Persian
  10. Punjabi
  11. Russian+Bashkir+Tatar
  12. Santali
  13. Spanish
  14. Swahili
  15. Tamil + Marathi
  16. Vietnamese

We also accumulated data on why people support or do not support, which garnered some broad insights.

Reasons Expressed for Support[edit]

All in all the facilitators received responses from 1466 Wikimedians, of which 60.7% users strongly supported the idea of drafting and enforcing a universal code of conduct in the movement. 20.8% users supported it on the condition that the drafting process must be thorough and transparent followed by a structured enforcement mechanism.

While more detail is available in each of the individual reports, Individuals who supported a universal code of conduct in aggregate expressed a hope that a UCoC will:

  • ensure diversity and inclusion in the movement,
  • give a safe platform to diverse voices and
  • ensure a collaborative working atmosphere in the movement.

Communities collectively stressed that harassment and discrimination have no place in the movement. Participating members reported that in many communities, even experienced editors and admins often cannot freely talk about illicit behaviour by some editors and the need for enforcing stronger conduct policies for the fear of community retributions and harassment. This prevented even community leaders from openly engaging in universal code of conduct discussions.

“Many Wikimedians have stopped contributing to the projects because of harassment/personal attacks. This needs to stop. Universal code of conduct is very important and necessary. It should have been prepared a long time ago.”

            - A long term Indic language community member

Community members appreciated the idea that smaller and upcoming communities will have the flexibility to build upon the UCoC to develop more comprehensive local policies. Users believe that a UCoC will not just give smaller and new communities a stronger sense of belonging in the movement but will also encourage them to actively participate in movement-wide discussions.

“A UCoC will help Wikimedians to feel free in contributing to WMF projects and will grant them to experience the freedom of expression especially when there is a difference in points of view.”

            - A Swahili language community member

Reasons Expressed for Concern[edit]

9.3% of the participants were opposed to the idea of a universal code of conduct. Along with raising legitimate apprehension about the complexity of creating universal values, these users expressed their concerns about safeguarding the autonomy of local communities. They feared that UCoC might pose a challenge to the tradition of the local governance system in communities. Some communities voiced that they do not have a healthy working atmosphere in the communities but they did not think a universal code of conduct would solve these challenges.

Other themes for opposition included a lack of transparency in the process, asynchronous communication, and inadequate faith in the foundation’s capability of forming a set of guidelines that takes into consideration diverse cultural nuances and sensitivities.  

“Cultural differences in the communities will make it difficult to have just one UCoC. It depends on the people.”

            - A Chinese language community member

“I can compare a universal code of conduct with the European Union law, which applies on all members, but it’s not permanent, where if one gap is detected, or if it ignores a certain group, the entire Union may collapse, England is an example. :)”

            - An Arabic language community member

Suggestions from neutral parties[edit]

Although 2.2% of the participants remained neutral and 7% did not take a clear stand on the subject, they did offer some key suggestions. Community members said that drafting the universal code of conduct will not be enough for it to be effective; it needs to be impartially enforced throughout the movement. They said that for it to be truly universal it needs to broaden its scope from communities and projects to virtual and in-person events as well as Wiki related discussions on other platforms. The code once drafted, should also be translated in multiple languages, periodically reviewed and if required, amended. A dedicated committee should be set up for this purpose.

“Harassment often tends to move from Wikipedia to social media websites, since people give more personal information on social media. A UCoC thus should apply to off-wiki platforms as well.”

            - A Vietnamese language community member

“Once formed, a universal code of conduct must be periodically reviewed.”

            - A Farsi language community member

All the supporting statements, concerns and suggestions shared by the community members have been duly noted and will be taken into consideration while drafting the code of conduct.

Collective feedback from all the communities consulted

Insights[edit]

The local language consultations were significantly insightful and helped us understand the distinct needs of the individual communities. The facilitators had constructive discussions even with communities that are not historically known to engage well with WMF. We heard many stories that are not often openly told in the movement. A glimpse of community-specific stories can be found in dedicated language summaries. Not all the stories could be published and even the ones that are presented there are highly anonymised to protect the privacy of the users. Some of the stories published reflect patterns we saw from several communities while some highlight very unique issues.

It was brought to our attention that there is a lack of awareness about even existing behavioural policies in many communities and not all of them have the time and resources to build comprehensive policies of their own.

We learned that different communities have different priorities and requests from the universal code of conduct. For example, one community more or less unanimously expressed its disappointment about the abuse of rights by admins and wants this issue to be addressed in a universal code of conduct. Another community stressed the need for special attention to religious sensitivities. A set of communities brought to the fore the pressing needs to address recurring harassment of female editors, and there was one community that went to an altogether different direction saying it does not encourage non-discrimination policies. The stories and demands that surfaced through the local language consultations are diverse and addressing them in the universal code of conduct will be a challenge, but we are committed to working with the communities to collectively address them.

How these consultations will be used and next steps[edit]

In July 2020, a Universal Code of Conduct Drafting Committee will start working on the actual draft language of the UCoC. The Committee will be taking into consideration the findings from these consultations, as well as upcoming conversations, to adapt a base text into a language that reflects the needs and concerns of Wikimedia communities.

A summary of the timeline for the UCoC project can be found on the main project page.

Footnotes[edit]

  1. The study suggested that of the 98 Wikipedias we studied, 52 (53% ) had little or no behavioural policies. 16 have similar names and content as English Wikipedia policies, 16 more have a combination of original and translated policies. And 14 projects that have at least one unique policy that takes into account their specific cultural context.
  2. While questions were asked around support for a UCoC, the intention is to understand factors that matter to these communities. This study is not meant to be a conclusive statistical representation of support or opposition for a UCoC.