[Wikimedia Brasil] The Missing Wikipedians

[Cecilia Tanaka]

The Missing Wikipedians <http://hblog.org/writing/the-missing-wikipedians/>

 The Missing Wikipedians

Download PDF
<http://makebuildplay.files.wordpress.com/2011/02/makmende_essay.pdf>or view
on Slideshare <http://www.slideshare.net/hfordsa/the-missing-wikipedians>

Much has been said of the future of Wikipedia. Some have prophesied that the
online encyclopaedia will fail due to increasing spam. Others have said
that, as large parts of the world go online, Wikipedia might see a wave of
new editors as countries from Zambia to Indonesia begin to fill in
Wikipedia’s blank spots. In a project that aims to ‘make all human knowledge
accessible’, those blank spots can mean many things: the hundreds of
thousands of places that aren’t talked about on Wikipedia, the thousands of
languages that either don’t have their own encyclopaedia or are struggling
to build one, and the countless things that people know about their world
but aren’t in written form.

This essay is concerned, not so much with the future of the English version
of Wikipedia (about which much of the prophesying occurs) but with the 277
other language Wikipedias. Will this number shrink as editors grow tired of
their lonely pursuits, or will it grow as more of the world goes online? As
large parts of Africa go online, it is expected that they will start to edit
Wikipedia and that they will edit it in their own language. Both of these
assumptions may be incorrect. Firstly, there are a number of external and
internal limitations to this new wave of editors joining Wikipedia, and
secondly, the scale of smaller Wikipedias may mean that they are
over-shadowed by stronger motivations to edit the larger, more powerful
English version.

*‘Makmende’s so huge, he can’t fit in
Wikipedia[2]<http://hblog.org/writing/the-missing-wikipedians/#_ftn2>’
*

In mid-2010, a furore erupted in a small corner of the Internet. The facts
sounded all-too familiar: another group of Wikipedia editors fighting over
whether something was notable or not. The so-called ‘deletionists’ against
the ‘inclusionists’ – those who thought that the encyclopaedia needed to
retain a certain quality and that strict editorial control was necessary,
versus those who thought that Wikipedia’s goal is to be a different
encyclopaedia – one that is much broader and more global than any other
existing encyclopaedia.

But a closer look at this blip on Wikipedia’s radar exposed some interesting
details – details that exposed this as a story that epitomises Wikipedia’s
current growth problems and the challenges it faces as it seeks to ‘make all
human knowledge accessible’. The frontline of this battle: a page called
‘Makmende’ that was struggling to be born on the English encyclopaedia.

In March of 2010, Kenya had enjoyed what has been touted as its first viral
Internet sensation. While even Eastern Europe has had its share of singing
kittens and political remixes, this East African country had not enjoyed the
success that comes when the world recognises a local meme that captures the
imagination of those outside of it. The meme was based on an interesting
local hack of Hollywood culture that originated on the streets of Kenya in
the 1990s.

The Swahili slang (sheng) word for ‘hero’, ‘Makmende’ originates from a
mispronunciation of Clint Eastwood’s phrase “Go ahead, make my day” (Mek ma
nday) – a phrase that became popular in the streets of Kenya in the 1990s
when a ‘bad guy wannabe would be called out and asked “Who do you think you
are? Makmende?”’ In early 2010, local band, ‘Just a Band’ resurrected the
fictional Kenyan superhero in the music video for their song *Ha-He*. In the
music video for their song, the band features Makmende beating up the ‘bad
guys’ and even ignoring the girl in a hilarious throwback to the fictional
character.

What followed was a popular acknowledgement of Makmende that resonated
outwards from local Twitter users. Like other successful memes, Makmende
enabled people to participate in the joke and to thereby “own” a little
piece of the meme. According to local digital marketing strategist, Mark
Kaigwa, people either took popular Chuck Norris jokes and replaced them with
Makmende, or they created their own. Radio stations in Nairobi invited
people to call in with Makmende jokes when local journalists like Larry
Madowo noticed the attention that Makmende was getting on Twitter, and the
Kenyan twittasphere seemed to be buzzing with their own Chuck Norris.

In the midst of enthusiasm, Makmende fans tried to create a Wikipedia page
about the meme. Wikipedia admins repeatedly deleted the page, initially on
‘criteria for speedy deletion’ G1 (‘Patent nonsense, meaningless, or
incomprehensible’), then G12 (‘Unambiguous copyright infringement)’ and
finally G3 (‘Pure Vandalism’).

Wikipedia editors claimed that the article needed to be deleted because
there existed ‘no reliable sources, and no claims of notability’. Pointing
to the lack of sources relating to African culture online, user, Cicinne
came back with this retort: ‘The problem is that there is hardly any content
on African influences in the 90′s and 80′s which may make it hard to make
the connections’.

On March 24, the Wall Street Journal’s Cassandra Vinograd commented on the
story, reporting that ‘Kenyan bloggers and Tweeters (had) seized on the
video and launched a campaign for the man they’re calling Kenya’s very own
Chuck Norris – complete with one liners about Makmende’s superhero skills
and prowess.’ According to the WSJ, Makmende had drawn more than 24,300 hits
in the week since its release and had collected 19,200 fans on Facebook.

The article was deleted once again, prompting Ethan Zuckerman to write a
blog post about the systemic bias operating in the encyclopaedia community
that would delete the stub:

The one that’s currently under development followed a classic Wikipedia
structure – it went up as a brief stub, and has accreted more content in the
past few hours. What concerned me is that the attempt to delete that stub
argued that the article was unsourced – actually, it was quite well sourced,
including a reference to a Wall Street Journal online publication and five
weblogs. Perhaps the user who nominated for deletion made a mistake. Or
perhaps he acted in bad faith, trying to avoid a battle over notability and
tried a different tactic to see the page removed.
If Wikipedia wants to make progress in improving areas where it’s weak –
i.e., if it wants to address issues of systemic bias – the community needs
to expand to include more Wikipedians from the developing world. Deleting
three versions of an article important to Kenyans and trying to delete a
fourth doesn’t send a strong message that Wikipedia is the open and
welcoming community you and I both want it to be.

After being covered on CNN, Fast Company and numerous location Kenyan
publications (most of which are not online), the article was eventually
voted ‘keep’ citing the WSJ post as proof of notability required to survive
and move past the deletion debates. The question then became: if something
needs to be ‘notable’ to get on Wikipedia, by whose standards are we judging
notability? Is it about numbers, about reputation? Can this be measured? And
would this have been such a debate if it had occurred elsewhere in the
world?

This story epitomises the challenges facing Wikipedia as it comes up against
the scope of a traditional encyclopaedia. Ethan Zuckerman summed it up as
follows:

Most Wikipedians seemed to accept the idea that different languages and
cultures might want to include different topics in their encyclopedias. But
what happens when we share a language but not a culture? Is there a point
where Makmende is sufficiently important to English-speaking Kenyans that he
merits a Wikipedia page even if most English-speakers couldn’t care less? Or
is there an implicit assumption that an English-language Wikipedia is
designed to enshrine landmarks of shared historical and cultural importance
to people who share a language? * *

Interestingly, Makmende does not exist in the Swahili version of Wikipedia,
and the battle to put Makmende on Wikipedia came just two months after
Kenyans were being incentivized by Google to create Swahili Wikipedia pages.
There seems to be a disconnect between where ordinary Kenyans want their
cultural narratives to live, and where outsiders imagine it.

This story doesn’t only represent a clash between the inclusionists and
deletionists in Wikipedia. It also reflects key issues about the
relationship between different Wikipedias in countries where English
dominates as the written language; about the motivations of Wikipedians on
the edges of the Wikipedia network; and about tensions between existing
policies, the goal of the encyclopaedia and the realities of historical
knowledge in the developing world.

*Background: Wikipedia growth is slowing *

In August of 2006, Diego Torquemada drew a statistical model that predicted
the future growth of English Wikipedia to reach 6 million articles by the
end of 2008. This model was based on the premise that more content leads to
more traffic which leads to more edits which generates more content on the
encyclopaedia. Wikipedia had enjoyed exponential growth until that point,
with the number of articles doubling annually from 2002 to 2006.

Torquemada could not know that Wikipedia growth had reached its peak in 2006
when he developed his model. At a rate of 60,000 articles per month in
mid-2006, the number of new articles would start to follow a downward trend
reaching the point of around 35,000 new articles per month by the end of
2009. The number of edits similarly reached a peak in 2007 with 6 million
edits and active editors at 800,000. At the end of 2009, the number of edits
had levelled out to about 5.5 million and active editors were down to around
700,000.

The slowing growth of Wikipedia has been the subject of a number of news
articles, as Internet commentators predict the slow demise of Wikipedia, and
Wikipedians fight back, saying that they are merely “consolidating”.

In trying to understand the slowing growth of Wikipedia, researchers at Palo
Alto Research Center took a closer look at the data and interpreted an
ecological model to explain the slowing growth. Suh, Convertino, Chi and
Pirolli likened the stagnation to a Darwinian ‘struggle for existence’ in
the encyclopaedia, noting that ‘as populations hit the limits of the
ecology, advantages go to members of the population that have competitive
dominance over others’.

Suh et al argued that the ‘resource limitations’ can be likened to limited
opportunities to make novel contributions and that the consequences of these
increasing limitations will manifest itself in increased patterns of
conflict and dominance. Wikipedians, it seemed, had covered all the “easy”
articles and now had “nothing left to talk about”.

*Nothing left to talk about? *

Is Wikipedia really ‘running out of things to talk about’? Suh et al
suggested that the number of Wikipedia articles could increase due to the
growth of new knowledge as a result of new scientific studies and new events
but that the size of the encyclopaedia was still levelling out.

Others like geographer, Mark Graham deride claims that Wikipedia is ‘running
out things to write about’ for other reasons. Mapping the presence of
geotags on Wikipedia, Graham found that there are still ‘whole continents
that remain a virtual “terra incognita”’ on Wikipedia and that if these
places were given the same detailed treatment as places in Western Europe
and North America, then Wikipedia is only just getting started.

*New Wikipedians as the developing world comes online?*

Graham suggests that, ‘It may be that when broadband reaches more parts of
Africa – helped by the landfall of superfast cables in August – that more
people there will start discovering Wikipedia, and that the site will see a
second explosion of new editors and articles about places that have so far
been ignored’.

But it is doubtful whether Internet access alone will make people in
developing countries contribute to Wikipedia. In his study of twelve
different Wikipedia language versions, Morten Rask found that although
‘there is a linear relation between the level of internet penetration and
reach of the Wikipedia network, there is a stronger linear relationship
between the level of human development and internet penetration’.

Rask used the United Nations Development Programme’s Human Development Index
in his study as a comparative measure of life expectancy, literacy,
education, and standard of living for countries worldwide. He was interested
in finding out whether Wikipedia was only for ‘rich countries’ in order to
understand ‘who is open to work together in the sharing of knowledge’.

Rask’s findings contradict the so-called ‘techno utopians’ who have claimed
that the mere existence of either the Internet or information and
communications technology have the ability to lift developing countries out
of poverty.  Techno utopians include commentators like Don Tapscott who
coined the phrase *wikinomics* to describe ‘deep changes in the structure
and modus operandi of the corporation and our economy, based on new
competitive principles such as openness, peering, sharing, and acting
globally’.

Tapscott believes that we are living through a ‘participation revolution
(that) opens up new possibilities for billions of people to play active
roles in their workplaces, communities, national democracies, and the global
economy at large. This has profound social benefits, including the
opportunity to make governments more accountable and lift millions of people
out of poverty’.

Access to Wikipedia’s ‘revolutionary’ potential is an extension of this
techno utopian vision. Investigating the ‘reach and richness’ of Wikipedia,
Rask provides a solid critique of statements like Tapscott’s that ‘all one
needs is a computer, a network connection, and a bright spark of initiative
and creativity to join in the economy’ by showing that ‘Internet penetration
is not the only complete and sufficient variable’ for development. Analysing
data from twelve Wikipedia language versions, and mapping it to variables
such as the country’s Human Development Index and broadband penetration,
Rask was able to show that human development variables were much more
critical to participation in Wikipedia than broadband access.

*Internal limitations*

Apart from the external limitations of human development and broadband
penetration, Wikipedians on the edges of the network also face a number of
internal challenges that reflect a growing resistance within Wikipedia to
new content. As those from developing countries come online and try to edit
the encyclopedia, a number of conflicts have arisen due to tensions between
so-called ‘inclusionists’ and ‘deletionists’ in the encyclopaedia.

‘Inclusionists’ are Wikipedians who would rather see more articles – even if
they are short and/or poorly written, while ‘deletionists’ are concerned
with quality, believing that it is more important to have less, good quality
articles than more poorly written articles with questionable notability.

In an article entitled, ‘The battle for Wikipedia’s soul’, The Economist
writes: ‘The behaviour of Wikipedia’s self-appointed deletionist guardians,
who excise anything that does not meet their standards, justifying their
actions with a blizzard of acronyms, is now known as “wiki-lawyering”’.

The Palo Alto Research Center group suggested that the ‘deletionists might
have won’ when they found that the number of reverted edits has increased
steadily, and that occasional editors experience a visibly greater
resistance compared to high-frequency editors.

According to Suh et al., ‘Since 2003, edits from occasional editors have
been reverted (at) a higher rate than edits from prolific editors.
Furthermore, this disparity of treatment of new edits from editors of
different classes has been widening steadily over the years at the expense
of low-frequency editors. We consider this as evidence of growing resistance
from the Wikipedia community to new content, especially when the edits come
from occasional editors’.

* *

*Public goods and the costs of contribution*

If Wikipedia is available in Swahili, and the effort required to start a
Swahili page is lower than on the English version, why was the Kenyan
community so determined that the Makmende article exist on the English
version of Wikipedia?

Clues to the answer can be found in debates about public goods. Wikipedia
can be considered to be a public good since it is non-rivalrous (one
person’s use of Wikipedia doesn’t deplete another person’s use of it) and
non-excludable (no one can be effectively excluded from using Wikipedia, if
they’re online at least). Peter Kollock, writing in the late 90s about
public goods and how their value shifts when it is placed online, declared
that all online community interaction creates public goods and that this is
a remarkable property of online interaction and unprecedented in the history
of human society.

Unprecedented as it is, people still need to be motivated to contribute to
public goods. The question with regard to the Makmende case is: If people
will create public goods when motivations are higher than costs of
contributing, what are the relative costs for contributing to English vs
Swahili Wikipedia?

It is clear from the Makmende example that Wikipedia newbies must navigate a
growing bureaucracy and complicated policies when dealing with English
Wikipedians, many of whom would rather not have to deal with any more
articles to improve. This creates a high barrier to entry that must be
offset by higher motivational factors in order to incentivise volunteer
activity.

If the costs of contribution in terms of centralised control, bureaucracy
and the lack of ‘reliable’ sources are higher in the English Wikipedia, then
motivations for contributing must have been significantly higher for Kenyans
when contributing Makmende to the English version.

In his paper on ‘The Economies of Online Cooperation’ Kollock notes four
motivations for providing public goods including anticipated reciprocity,
reputation, sense of efficacy and need.

*Reciprocity*

According to Kollock, ‘a person is motivated to contribute valuable
information to the group in the expectation that one will receive useful
help and information in return that is, the motivation is an anticipated
reciprocity’.

The promise of reciprocity on the English Wikipedia is relatively high based
on the scale of contribution. Even though contributors account for less than
1% of users, the scale of the encyclopaedia means that the numbers of active
contributors is about 40,000 active editors for 26 per million speakers
versus Swahili Wikipedia with 0.4 editors per million speakers (about 20
active editors). According to Phares Kariuki, he created the Makmende page
because there are few opportunities to create a Wikipedia entry that would
be populated quickly. Kariuki said that he isn’t a regular Wikipedia
contributor and that the last time he contributed was many years ago. He
points to the small numbers who care enough to promote the page as a
problem. “If I started a page on my high school it would take six years to
build up.” Kariuki had tried to edit before but didn’t have much success. “I
am a heavy user like most of us here in Nairobi but there’s never really
been motivation to become an editor before,” he said.

Wikipedians on the English Wikipedia are relatively assured that others will
continue to contribute, whereas contributors to smaller Wikipedias must
understand that numbers of editors are few and that Wikipedia may shut down
Wikipedias where growth has stagnated and where they have become overrun by
spam.

Interestingly, Eric Goldman’s claim that ‘Wikipedia will fail in 5 years’
because of increasing spam has been more prophetic for smaller Wikipedias
than the English Wikipedia. According to Goldman, ‘free editability’
(allowing anyone to edit) is Wikipedia’s Achilles’ heel. The sheer scale of
the English Wikipedia has won out against spammers in English Wikipedia, but
smaller Wikipedias must face a continual battle – especially when their
numbers are so small in comparison to the spammers.

*Reputation*

Kollock noted that the effect of contributions on one’s reputation is
another possible motivation. ‘High quality information, impressive technical
details in one’s answers, a willingness to help others, and elegant writing
can al work to increase one’s prestige in the community,’ he found.

It is interesting to note that the reputation motivation requires that there
are people to impress in the community. Because of the small scale of
Swahili Wikipedia, for example, the fact that one can gain prestige from the
group might not necessarily be positive if the real power lies outside the
group. The English version of Wikipedia receives 9 million views per hour,
whereas the Swahili version gets 1,700 with the effect that one’s reputation
is much more highly valued on the English version of Wikipedia.

In addition, the content of the article is noteworthy. A description of
Kenya’s first Internet meme, it can be seen as Kenya’s unique contribution
to the global phenomenon of Internet memes. This wasn’t an article about the
British parliamentary system or the life cycle of bees – it was an article
that positioned itself in the global framework of Internet memes. ‘Look,
world,’ Kenyans seemed to be saying, ‘You have your Internet memes. Now we
do too!’

If one looks at this through the information sharing lens, one can make a
parallel with the fact that people are more likely to contribute expertise
rather than organisational knowledge because of its unique character and
because it shows something of their unique nature. Kenyans were sharing this
information specifically on the English Wikipedia because it was unique in
the global sense and because they were about to contribute their expertise
on a subject that they had direct experience with for the first time.

*Sense of efficacy*

The third possible motivation proposed by Kollock is the sense that a person
contributes valuable information because the act results in a sense of
efficacy, that is, ‘a sense that she has some effect on this environment’.

Certainly, those editing Swahili Wikipedia must have a much larger sense
that they are affecting change in the environment since their edits are much
more likely to be accepted, and they are more likely able to develop
policies and rules in the emerging Wikipedia. Contrast this with the fact
that new content on English Wikipedia will most likely be reverted and one
recognises how one’s sense of efficacy on the environment is affected by
Wikipedia’s growing isolation from new editors.