• Why Do People Write for Wikipedia?

Sponsored Links

  • FileName: ForteBruckmanWhyPeopleWrite.pdf [read-online]
    • Abstract: Why Do People Write for Wikipedia?Incentives to Contribute to Open-ContentPublishingAndrea Forte and Amy BruckmanGeorgia Institute of Technology, College of [email protected], [email protected]

Download the ebook

Why Do People Write for Wikipedia?
Incentives to Contribute to Open-Content
Andrea Forte and Amy Bruckman
Georgia Institute of Technology, College of Computing
[email protected], [email protected]
Abstract. When people learn that we have spoken to individuals who spend up to 30 hours a
week volunteering their time to research and write for an open-content encyclopedia, we often
hear the same question: “Why do they do it?“ The fact that this encyclopedia does not provide
bylines to credit authors for their hard work makes the scenario still less fathomable. Two rounds
of interviews with 22 volunteer encyclopedia writers in the fall of 2004 and spring of 2005
revealed that, in some respects, the incentive system that motivates contributions to the open-
content encyclopedia Wikipedia resembles that of the scientific community. Like scientists,
contributors to Wikipedia seek to collaboratively identify and publish true facts about the world.
Research on the sociology of science provides a useful touchstone for considering the incentive
systems embedded in the technology and culture of online communities of collaborative
authorship. In this paper we describe some of our findings in the context of Latour and Woolgar’s
seminal work on the incentive systems that motivate publishing scientists. We suggest that
minimizing reliance on “hard coded,“ stratified user privileges and providing indicators of
engagement in desirable activities can help support the growth of incentive economies in online
Prolific communities like Wikipedia provide interesting cases for examining technical
and social incentive mechanisms in online communities. Wikipedia is a highly productive
community of collaborative authorship in which individuals contribute to an open-content
encyclopedia by adding content and revising others’ work online. Like scientific
communities, contributors to Wikipedia seek to collaboratively identify and publish true
facts about the world. Unlike scientific publications, Wikipedia does not explicitly credit
authors for their work in bylines; the radical collaborative model simply does not afford
direct attribution of authorship. By using traditional communities of authorship and
publication as a point of departure, we can ask interesting questions about the nature of
incentives in Wikipedia and similar online communities. In what ways does the incentive
structure that supports the process of publication and fact-making dialog in the sciences
reveal itself in the Wikipedia community? What are the implicit messages about
collaboration and truth that are embedded in traditional incentive systems associated with
publication? Do the social and historical meanings of publication as incentive in the
sciences lend themselves to replication in a radically collaborative authoring environment
like a wiki?
We seek to develop an understanding of open-content publishing systems by first
describing research on incentive systems in the scientific publishing community and then
using this model as a framework for understanding the experiences of contributors to the
Incentive, Authorship and the Scientific Community
Before we can understand the ways that the long-standing practices of the scientific
community can guide our investigations of online communities, we must identify salient
features of incentive in the sciences. Among the best-known and earliest sociological
explorations of scientific practice were Latour and Woolgar’s anthropological
investigations of scientific laboratories (Latour & Woolgar, 1986). Alongside a wealth of
information about how scientific practice is situated in its instruments, offices, and
laboratory spaces, these studies clearly described an elaborate incentive structure linked
to publication.
Latour and Woolgar found that, for the scientific community, the most critical organ
of the incentive system is the cycle of credit. Credit refers simultaneously to two
dimensions of social status in the scientific community. First, it is fundamentally linked
with an individual’s ability to act in the community and effect change through the
assertion of claims. It is in this first sense that the cycle of credit describes how credit
becomes manifest in grants, equipment, data, ideas and publications. Second (and
secondarily), credit is a reward mechanism that marks one’s past contributions. The
notion of credit is complex and accruing credit is no straightforward process; it is
associated with a continuous social process of peer-review and allocation of prestige over
long periods of time. Early in their work, Latour and Woolgar observed that scientists
often used the term “credit” to describe something given or received; however, in the
course of their studies, they realized that the term stood for much more than the metric of
a straightforward reward system. In fact, what seemed to drive scientists and motivate
them was a sense of credibility that allowed them to assume more and more central roles
in the scientific community. In its fullest sense, “credit” is not something that is given or
received by individuals in the community, but a measure of power and efficacy.
It is important to note that although credit is linked with publication, the relationship
between accruing credibility and publishing is subtler than the simple act of claiming
authorship. Instead, it is a continuous cycle of strategic resource reinvestment of which
publication is only one part. This will be important as we consider the incentive structure
in Wikipedia. In summary, Latour and Woolgar found that reward is not the basis of the
incentive system in the scientific publishing community; instead, incentive involves the
allocation of power, resources, and the ability to reinvest and realize the benefits of
reinvestment (in the form of more credibility) as quickly as possible.
The Case of Wikipedia: Interviews with 22 Wikipedians
Two rounds of interviews with 9 and 13 Wikipedians in the fall of 2004 and spring of
2005 revealed that in some respects, the incentive system that motivates contributions to
Wikipedia resembles the incentive system observed in the scientific community. The
notion of credit exists in Wikipedia both as reward and as credibility that empowers
individuals in the community. Still, the nature of the encyclopedia-writing enterprise, the
technology on which the community is built, and the values of the community change the
incentive system in important ways.
Perhaps the most flagrant difference between the scientific community and Wikipedia
is the indirect attribution of authorship. On the surface, it appears that contributors
receive no credit whatsoever for their contributions. None of the articles are signed; most
have been edited numerous times by numerous people and explicit attribution would
seem to be impossible. In fact, interviews revealed that Wikipedia authors recognize one
another and often claim ownership of articles:
At this point, about 1600 English language articles [are] on my watchlist. Some of those are my
own, some of those I’ve made significant contributions to but didn’t start them, and a good number
of those—because I’ve become a well-known person in that community—somebody has
specifically invited me to come in and look at the article because it’s one that has had controversies.
For example, a lot of articles related to the Israel-Palestine situation, I’ve done very little writing on
them myself, but I have been very involved in the discussions on it. (BJ-1)
In some ways you get recognized, you get some respect, recognition from your fellow…here’s
somebody who knows his stuff, who writes good articles and so on and so forth, and you feel happy
when one of them puts a posting on your talk page. (SB-5)
Recently I’ve been working on the article… as a featured article candidate. If my article is accepted
as a featured article, it will appear on the main page with a multi-paragraph excerpt and photo.
Featured articles stay on the front page for a day, and then they’re swapped for another, so I’m
really just trying for bragging rights with this one. (SB-9)
One of the things I try to do if I write a new article, I try to get it put on the front, on the main page,
there’s a section called “Did you know?” …Once it gets there, it usually attracts 10 to 20 people in
the next few days. They’ll come and not only work on it but put it on their watchlist, when that
happens it definitely makes the article much, much better. (JM-1)
Although the culture forbids individuals from explicitly claiming authorship within an
article—it simply isn’t done—the technology provides indirect ways of establishing
ownership. First, the editing history is available for every article, so it is possible to
ascertain who created an article in the first place and to review each change to identify
the most substantial contributions. Second, contributors often claim ownership of articles
on their own user pages by creating lists of the articles for which they believe they ought
to receive credit. Some Wikipedians include elaborate “resumes” on their userpages.
We see both through observations on the site and participants’ descriptions that
authors claim and receive credit in Latour and Woolgar’s secondary sense—as reward for
a contribution to the community. How does credit manifest itself in Wikipedia as
credibility that can be reinvested in the encyclopedia-building endeavor?
There are both technological and human resources that administrators use to influence
the character of Wikipedia content. By what processes is credit invested in these
resources? Anyone may become an administrator; however, since administrators are
voted in, one must have accrued the requisite credibility in the community to receive
unanimously positive votes.
Accruing credit is not the process of a straightforward meritocracy. One Wikipedian
expressed his frustration with the politics of credit on Wikipedia:
You have… people who never ever get name recognition at all, but they’ve created a huge amount
of high quality content and haven’t caused trouble and have behaved themselves and nobody knows
them. For example, when I was running for administrator a couple of votes said “I never heard of
this guy” about me, even though I had done quite a bit of work. Then I see other people getting
voted to be an administrator and everybody is making the comments “Gosh you’re not an
administrator already? I thought you were already administrator. Of course I’m gonna vote for you”.
I compare our relative edit history and this other person who everybody knows really has not created
that much content. (JM-2)
Aside from confusion about how it is appropriate to make a name for oneself, there is
controversy over whether one should make a name for oneself. Many members of the
Wikipedia community subscribe to a populist, egalitarian view of knowledge production.
This stands in contrast to our comparison case, the scientific community. Reconciling an
epistemological commitment to the experiences of the everyman with a need to create an
encyclopedia that is perceived as being a reliable resource has led to confusion about the
role of identity and one’s credentials. Some Wikipedians expressed indifference to the
possibility of establishing an identity on the site; clearly, the relationship between
individual contributors’ authority and the reliability of the Encyclopedia is not well
One only need ask a librarian about the reliability of Wikipedia content to discover
that credibility in Wikipedia is a localized phenomenon. Credibility in the Wikipedia
does not translate to credibility in broader social contexts in the same way that credibility
in the sciences does, nor is the ability to create content and effect change in the
Wikipedia community firmly moored to credibility as in the scientific community.
Although local mechanisms have emerged for accruing credit, it is easy for a newcomer
to enter into the community and quickly begin to direct the character of the content and
the discussions that happen on the site. Without this relatively low barrier to entry, the
community would be difficult to sustain. In fact, a more rigorous peer-review process
resulted in the demise of the Nupedia, which was the original form of Wikipedia (Sanger,
2005). One interviewee expressed a concern that credibility on Wikipedia and credibility
in broader contexts could be conflated:
I’m not happy with the way that you can start a page for yourself or your username. That doesn’t
seem like it makes much sense to me. That seems like people could be using it for nefarious
purposes, which I wouldn’t doubt. And people can also make themselves look credible through what
they post on Wikipedia. (AF-1)
Despite the many similarities we observed, one final, important difference sets
Wikipedia apart from the scientific community. Wikipedians are not engaged in
primary research. The encyclopedia-writing endeavor requires a different kind of
credibility than scientific inquiry. For encyclopedists, it is important write well from
multiple sources and reliably assess those sources. These are the activities to which
credibility in Wikipedia must be linked in order to sustain productive investments
from members.
Implications for Designing Online Communities
Naturally, an economic model of incentive fails to account for many facets of motivation
that attract and sustain involvement in communities, scientific or otherwise. Still,
whatever the motivations that participants bring to a social interaction, an incentive
economy provides a continuous and coherent framework for sustaining participation.
Whether a scientist first enters a field to help ensure that her country has clean drinking
water, because she feels compelled to fulfill familial expectations, or because she had a
fantastic biology teacher, the accumulation and investment of credit provides an incentive
system that lends structure to each contribution she will make throughout her career.
Likewise, the design of communities in an online environment must meaningfully
structure participants’ contributions in a way that sustains involvement. That means
moving beyond the use of straightforward rewards and into the realm of incentive
economies that allow productive participants to achieve higher levels of efficacy and
responsibility in the community. Amy Jo Kim’s recommendation to reward regulars and
empower leaders speaks directly to this need for incentive economies (Kim, 2000).
Latour and Woolgar’s model suggests that incentive systems need not result in explicit
stratification where some users fill leadership roles and others do not, but instead, can
lead to a system in which gradations of credibility (and power) are subtly determined
through interaction and discourse. This is accomplished by creating a need for
participants to invest themselves in the community. In Wikipedia, the power of
participants is not clearly circumscribed by the technology; although administrative
powers are held by some and not others, the process for granting and gaining
administrator status is open to anyone “off the street” who can provide a compelling
image of himself. We saw in interviews that many long-standing community members
and relative newcomers held equivalent technological power in the community. Yet,
long-standing members described activities that revealed their clout, such as liaising
between Wikipedia and other communities and establishing joint projects.
One way that members invested more of themselves in the community of Wikipedia
was through their presence on multiple channels of discourse. Discussion pages, meta-
pages, announcement pages, mailing lists, and IRC provide multiple channels for
involvement in the Wikipedia community. If a participant wants to accrue more
credibility, one way to do so is by participating in multiple channels; this requires a
substantial time commitment. The system works remarkably well as it is and we do not
fully understand the balance of the credit cycle in Wikipedia. One can imagine
experiments that explicitly anchor credibility in some desirable form of investment (say,
authoring rather than politics). For example, the software might tie some obvious
indicator of community members’ contributions to their username rather than relying on
participants to ferret out and interpret editing histories.
We suggest that very simple features in online communities can help guide the
development of an incentive economy. First, although some thresholds for granting
technical powers may be necessary, pains should be taken to keep “hard coded”
stratification of participants’ power to a minimum in order to allow for incremental and
socially-agreed upon leaders to emerge. Second, technology can be used to identify
engagement in behavior that we want to link with credibility. Unlike scientific
communities, new online communities do not have the benefit of hundreds of years of
refined methods or citation practices to help structure cycles of credit. Providing some
obvious indication of what kinds of activities warrant higher levels of credibility may
help fledgling communities grow in desirable directions as participants invest their time
and identities.
The Wikipedia studies would not have been possible without the interviewing skills of
Bill Julyan, Juan Munoz, Adam Wilson and especially Susan Bryant.
Kim, A. J. (2000). Community Building on the Web : Secret Strategies for Successful
Online Communities. Berkeley, CA: Peachpit Press.
Latour, B. & Woolgar, S. (1986). Laboratory Life: the Construction of Scientific Facts.
Princeton, NJ: Princeton University Press.
Sanger, L. (2005). The early history of Nupedia and Wikipedia: a memoir. In DiBona, C.,
Stone, M. and Cooper, D. (Eds.), Open Sources 2.0 Sebastopol, CA: O'Reilly Press.

Use: 0.6969