Wikimedia Blog/Drafts/The value of oral citations

From Meta, a Wikimedia project coordination wiki

Title ideas[edit]

  • Oral knowledge for Wikipedia
  • Tapping into the knowledge of indigenous communities
  • Wikipedia needs oral history
  • ...
Tatekulu Hijapendje pondering what to answer

Wikipedia has made tremendous progress towards its mission to provide free access to the sum of human knowledge, but indigenous knowledge is largely excluded because the majority of it is not available in writing. Starting from some theoretical considerations, I designed a workshop for the 2014 Participatory Design conference in w:Windhoek to produce and document examples of relevant oral citations. A detailed workshop description is available here.

In October a small group of Wikipedians traveled to the Namibian village of w:Otjinene to interview elders. The aim of the interview was to directly convert narratives into Wikipedia content with oral citations. We visited the homestead of Tatekulu Festus Hijapendje and his wife Memekulu Olga Muhaindjumba Hijapendje and asked them about traditions, culture and development of the local Herero community. First results are available at Wikipedia:Oral citations experiment/Articles.

Thanks to a Project and Event Grant from the WMF I could invite editors from the region to collect the narratives. Editors wishing to participate had to answer the Call for participation, adopt an article from a list of offers, and improve the article with conventional (written) references. In the village we then planned to ask questions around the 'blind spots' in the adopted articles, missing information that no written source could help fill. Now we can present two scenarios for a small set of Wikipedia articles: One restricted to ordinary, written sources, and one that utilizes narratives originating from indigenous knowledge. We hope to be able to dismiss the suspicion by Wikipedia's editor community that the online encyclopedia has nothing to gain from the inclusion of indigenous knowledge.

Challenges[edit]

The number of applicants was much lower than I had hoped for. From a planned group of twelve participants I could admit only four, one of which cancelled on short notice, and one editor from Ghana could not come because the Namibian authorities did not grant a visa, possibly due to the ongoing Ebola scare.

Down to a group of five, (two participants, the driver and translator, a filmmaker, and me) we traveled 250km into the w:Omaheke Region to collect knowledge from the oral repository of the w:Herero people. On arrival we found that a funeral of an important community member had all but emptied the village on this particular weekend. Only children, the elderly, and some very few inhabitants were around. On top of that, our original plan to attract interviewees by means of a free barbecue was ill-considered, few people had noticed our arrival, and the Hijapendje couple would not have been able to walk over to our accommodation.

Workshop participants in the elders' homestead. From left: Muhaindjumba Hijapendje, Gereon Koch Kapuire, Festus Hijapendje. From right: Bobby Shabangu, Peter Gallert


Convincing Wikipedia editors of the encyclopedic value of oral knowledge repositories has also so far been an uphill battle. This value needs to be discussed in two dimensions. First, the usefulness of narrated content for Wikipedia needs to be established, and second, 'good' oral citations need to be distinguished from 'bad' ones - as with written sources, by far not everything that could be cited, should be cited.


Gathering oral citations is a learning process, and we are just at the beginning. Because interviewers, literally, do not know what information they want, questions can never be very specific. And even if they are, the answer might not be, for example:

Q: Please tell us when this settlement was founded.
A: (after long deliberation) Many, many years ago.

The resulting narrative might take its own direction, and every once in a while needs to be steered back on course, of course without interrupting the elder. And finally, without a bit of prior insider knowledge about the indigenous community certain answers are impossible to understand, and certain subtleties cannot be captured:

Q: In a Herero family, who is making the monetary decisions?
A: (by the woman) We reach an agreement but the man has the final word.
Q: (out of a suspicion because the man did not say anything) So the woman is in charge of the household but the man is in charge of the money?
A: (by the woman) Yes.
Q: Imagine you only have 500 dollars, from which the school fee could be paid, or from which the car could be repaired. What will be the likely outcome?
A: (by the woman) The man will not know we have 500 dollars. I will have paid the school fees already.


Local community members are much less likely to misinterpret oral information than outside researchers; eventually such interviewing should be done by members of the community. But for now, it requires a great deal of Wikipedia knowledge and experience to make an oral citation stick. Experienced editors need to get the ball rolling, both by collecting oral citations and by participating in the inevitable policy debate. We now understand a bit better of how to collect oral citations, and how to select them for building an encyclopedia. WMF allowing, I will be soon in the village again to gather more.

Peter Gallert, Polytechnic of Namibia