Research:Indigenous Knowledge

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search
Maja van der Velden
Heike Winschiers-Theophilus

This page documents a research project in progress.
Information may be incomplete and change as the project progresses.

Key Personnel[edit]

  • Peter Gallert, Namibia University of Science and Technology
  • Maja van der Velden, University of Oslo
  • Heike Winschiers-Theophilus, Namibia University of Science and Technology

Project Summary[edit]

This project is a follow-up of Research:Oral citations, exploring different solutions for including indigenous knowledge in Wikipedia. We pay particular attention to not infringing on the intellectual property rights of the indigenous community. The project belongs to the intersection of the research cluster "Community Centered Localisation" at Namibia University of Science and Technology, and the "Namibia Knowledge Portal", a public-private partnership to supplement and improve education in Namibian schools.


  1. In its first phase this is theoretical work, analysing inside definitions of Wikipedia terminology and showing where they contradict established scientific results. This phases' methodology is synthesis.Yes check.svg Done
  2. In a second stage we will initiate policy discussions about the treatment of indigenous knowledge on English Wikipedia. We assume this to be the hardest part of the project, possibly running in parallel to all subsequent stages below.Yes check.svg Done
  3. The third stage will consist of two experiments:
    1. Using experienced en-wp editors (ideally those who will not make any mistakes other than using oral citations) we present a before-after scenario for a few articles where little written reliable sources can be found. We invoked this stage even though lack of progress in phase 2 did not give us much hope that the resulting changes to en-wp articles would have survived. The results are available on w:Wikipedia:Oral citations experiment and subpages.Yes check.svg Done
    2. Depending on how much leeway stage two will give us for an experiment, we might utilise some small language editions, all currently in incubator, to collect and orally reference indigenous knowledge. This is done in order not to endanger timely progress of the overall project, the preservation of indigenous knowledge as such. If we are forced to enter this stage we will also consider alternative places to store IK. This methodology would be a long-term experiment.Symbol wait.svg Doing...
  4. In order to measure whether our interventions result in any editing uptake on the African continent we want to measure the 2017 baseline per indigenous group. Detailed reasoning see below.

Detailed methodology for step 4[edit]

As we are in the process of applying for access to non-public data for #4 of the project plan above, let me outline rationale and methodology a bit:


Wikipedia usage on the African continent is much lower than elsewhere. This applies to both read access and editing. The standard explanation for this situation is poverty, illiteracy, and lack of infrastructure, and these surely explain some of this difference. However, our research so far points to various cultural differences as well, predominantly values, work organisation, and the position of the individual in the society. Wikipedia, through its design and organisation, encorporates values that indigenous communities in Africa do not share. One worthwile task can thus be to customize African language editions in a way that is better suited to the local population (for a list of results see User:Pgallert).

The participation of Africans is probably still far less than what can be estimated by the geo-location of edits. When African editors gather in person they form a group that, let me put it this way, is disproportionally white, compared to African population figures. This could, of course, have many reasons: access to infrastructure, willingness to apply for scholarships, self-confidence, and many others. But it could again point at differences in values and culture, and strengthen the point for local redesigns.

For now I wish to measure a baseline of editing by indigenous Africans. The evidence for my claim that much of the editing from the African continent is done by people with non-African descent---expats, missionaries, tourists, scientists---is so far only anecdotal. Many editors disclose publicly what their native language is but few state where they currently live and work.

While edits geo-locating to Africa are available from the publicly available dataset I wish to classify editors into four groups:

  1. Indigenous to Africa, residing in Africa
  2. Indigenous to Africa, not residing in Africa
  3. Not indigenous to Africa, residing in Africa
  4. Not indigenous to Africa, not residing in Africa

A detailed definition of "indigenous" that will be applicable to the African situation has in my opinion yet to be developed and will form part of later theoretical deliberations. The white settlers in Southern Africa, for instance, clearly have adapted to their environment and have developed knowledge that must be regarded as indigenous. The term indigenous can thus not just apply to non-white inhabitants of the continent.

The rationale for wanting to separate edit count per indigenous group is that our preliminary theoretical investigations point to cultural, rather than infrastructural, hurdles for African editing. The working hypothesis is that those cultural differences are less prevalent in the groups that descend from European settlers, in the group of recent (i.e., first-generation) immigrants to Africa, and among Africans in the diaspora.

What I want to do is[edit]

  1. Identify (very) active editors whose edits geo-locate to the African continent, down to a certain threshold (100 edits a month) but independent of the Wikimedia project,
  2. Determine if they are indigenous through:
    • self-disclosed 'native language' entries in their Babel-boxes
    • narrative and other clues on their user and user talk pages (e.g. alma mater, self-descriptions, pictures)
  3. Quantify how much African editing is actually done by indigenous people
  4. Determine who (if not predominantly African indigenes) edits African topics: Expatriates, members of the diaspora, or 'ordinary' editors from elsewhere

The expected result will likely not be a straight percentage, but rater a scale of certainty. We expect many ambiguous results, e.g. a fr-N editor from Gabon, who could be native African or a francophone expat. But spot checks show that often the membership to one of the classes [indigenous Africans|others] can be directly determined from the user page. The geo-location would enable me to split the first group further into [Indigenous African in Africa|member of the African diaspora].


  • That there are disproportionally few indigenous Africans among the group of editors that reside in Africa (compared to Germans residing in Germany or Russians residing in Russia, for instance)
  • That editors not editing from the African continent but self-identifying as native in an African language, are part of the diaspora
  • That active editors are likely to self-disclose the desired parameters, directly or indirectly

Assumptions that might be problematic and thus need further elaboration:

  • That all Wikimedia projects present roughly the same cultural values
  • That all indigenous Africans (even just within one country) share roughly the same cultural values
  • That the selection bias of picking self-disclosing editors does not invalidate the result
  • That the technical difficulties of disclosing a rare native language per {{babel}} does not invalidate the result: Everyone knows, or can advise on, {{lang-en}} whereas a userbox disclosing {{lang-naq}} is much lesser known or might even have to be created first.
  • That indigenous Africans would not self-identity as 'native' in their former colonial language. That is true for much of anglophone Sub-Saharan Africa with the possible exception of Zimbabwe. For lusophone or francophone countries I am not so sure at the moment. Afrikaans will be another special case, as it is the only language that both colonialists and indigenes really share.

The data we need[edit]

Of all editors in all WM projects, per month in 2017:

  1. whose IP address geolocates to Africa and
  2. whose global edit count reached 100 in that month

I need

  1. The account name
  2. The home wiki
  3. The country


Some results will be disseminated openly, some probably won't. We attempt to have all WikiMedia-relevant results available under CC-BY-SA, but the main researcher is also constrained by the regulations surrounding his PhD work. Some of the already published results are under copyright.

We will summarise our findings and attempt to publish them on Commons and/or on the WMF Global Blog. The main investigator is by now a frequent guest at Wikimania---for as long as the WMF gives him a scholarship he will present and discuss the progress of the project there.

Wikimedia Policies, Ethics, and Human Subjects Protection[edit]

The project is part of a research cluster whose activities are discussed and approved twice a year by the Institutional Research and Publication Committee of the Namibia University of Science and Technology, which also handles ethical implications. Several other ethical committees have approved the parent project for the purpose of establishing a cooperation agreement (CSIR Meraka Institute, University of Cape Town) or grant applications (UNESCO, private companies).

The project has been established precisely because of the negative impact of publishing and utilising indigenous knowledge, on the intellectual property rights of indigenous people, and all researchers are regularly briefed on ethical implications of their work.

For the work on non-public data, results will be summarised and anonymised in a way that the retrospective identification of individuals ist not possible. Due to the (assumed) small data base we plan to present the results per region (e.g. anglophone Southern Africa) rather than by country. The presentation of findings will be discussed internally before the results are published.

Benefits for the Wikimedia community[edit]

The project attempts to make indigenous knowledge acceptable for Wikipedia. If successful, it will bring WP a substantive step closer to presenting the sum of all human knowledge. The "formal outcome" (academic publications) will help supporters of this initiative formulate arguments to convince sceptics. Specific benefits include:

  • Of our outreach activities: Establish a readership for Wikipedia in rural Namibia, and equip rural community members with skills and knowledge to edit Wikipedia
  • Of our analysis of policies and guidelines on the English Wikipedia and their application: Improve accuracy and consistency of both the wording and the interpretation of en-wp's rules, and reduce the influence of systemic bias on encyclopedic content of Wikipedia
  • Of our involvement in policy discussions around IK inclusion: Provide a currently underrepresented view point of people who have experienced and researched oral knowledge transfer in remote communities
  • Of the proposed long-term experiment of collecting IK in one or two small incubator editions: Demonstrate that the body of IK is stable, consistent, and comprehensive
  • Of the proposed short-term experiment (workshop) of expanding selected articles with orally sourced information: Demonstrate what the inclusion of oral citations can do to improve Wikipedia

By concentrating on the English Wikipedia with its well-known scepticism of oral citations we hope that if we are successful, a change of regulations will more or less smoothly reach smaller language editions as well. Except for our planned long-term experiment we deliberately do not try to hide our intentions in the Incubator, or in some obscure local language wikis.


  • August 2013: Pre-release of the book chapter. Yes check.svg Done Preliminary title: Reliable Sources for Indigenous Knowledge: Dissecting Wikipedia's Catch-22 (355KB)
  • 11 August 2013: Present key findings at Wikimania 2013. Yes check.svg Done Title: Indigenous knowledge for Wikipedia -- Bending the rules? (4.4MB)
  • November 2013 March 2014 delayed, too busy: Book chapter submission due
  • Mid 2014: Indigenous Knowledge for Wikipedia workshop proposal submitted to the 2014 Participatory Design Conference. The proposal suggests to hold a workshop with experienced en-wp editors who will create a before-after scenario of articles where classic written sources do not exist. Yes check.svg Done The workshop was accepted into the conference program, the call for position papers was written and workshop participants were selected.
  • 14-16 October 2014: PDC Conference in Windhoek, Namibia. Our workshop was held as a post-conference event. Yes check.svg Done See WMF blog entry.
  • April 2014 End 2014? Mid 2015 delayed, publisher jumped ship: Book to be published
  • End 2015 delayed, organisational hiccups Early 2016: Indigenous Knowledge for Wikipedia village week as a planned follow-up on the IK workshop.
  • Early 2016: Creation of a few audio-visual content pages on OtjiHerero Incubator and testing of an app that simplifies the creation of image maps. In one of our ordinary outreach sessions I plan to present the OtjiHerero Wikipedia as audio-visual, starting with an image map and linking to videos, audio files, and further images.


  • The research cluster activity, including this subproject, is supported by Namibia University of Science and Technology, CSIR Meraka Institute (South Africa), Telecom Namibia, and many others.
  • The Namibia Knowledge Portal is supported by the Namibian Ministry of Education and Telecom Namibia.
  • The Wikipedia IK editing workshop is partly supported by the chairs of the Participatory Design Conference
  • For the Wikipedia IK editing workshop we have received a WMF grant: documentation.


External links[edit]