printable banner

U.S. Department of State - Great Seal

U.S. Department of State

Diplomacy in Action

Wikipedia in a Post-Fact World: Reliable Sources, Transparency, and Open Knowledge

Katherine Maher, Executive Director of the Wikimedia Foundation
Washington, DC
April 11, 2017


MODERATOR: All right, folks, so we’ve got three people in the room here in Washington. My name is Mark Zimmer; I’m a Media Relations Officer here at the Foreign Press Center. We’re very pleased to have with us Katherine Maher, who is the Executive Director of the Wikimedia Foundation. She’s joining us today to speak about Wikipedia in a post-fact world – reliable sources, transparency, and open knowledge. She’ll make some opening remarks, we’ll open it up to questions, and we’ll ask our colleagues in New York to make themselves known as they have questions, given that we’re on a voice conference, as opposed to a video.

With that, Katherine, thanks for joining us, and we appreciate you being here.

MS MAHER: Thank you. Thank you very much. And I am going to, as Mark said, make some remarks just about Wikipedia. Forgive me if this feels a little rudimentary; Wikipedia is well known by many people, but often not necessarily well understood. And so my goal is to give you just a little bit of context about how we work, which may then provide some insight into why we’re here, to talk a little bit about transparency and the creation of knowledge and the importance of accountability to our public.

So Wikipedia has been around now for 16 years. We were first created in January, 2001. The Wikimedia Foundation, which operates Wikipedia as a not-for-profit 501(c)(3) organization, has been around for 14 years now. Wikipedia has been built by volunteers over time. We have millions of contributors. We are now existing in about 300 languages with 40 million articles across those languages. There have been 3 billion edits to Wikipedia since it was first created, Wikipedia and the other Wikimedia free-knowledge projects. Wikipedia is edited about 350 times every minute, and we receive more than 200,000 people who contribute on a monthly basis. And those people are – many of them are the same every month, but many people stop by, make an edit, move on, and will contribute to us again when they see the next opportunity. Over the course of a given month, we receive visits from about a billion unique devices from all over the globe, making us one of the world’s most popular and used websites, one of the largest.

We have a vision at Wikimedia that animates the work that we do. Our vision statement is a world in which every single human can freely share in the sum of all knowledge. This is quite critical to how we understand our perspective in the world. As a non-commercial entity, we think of ourselves more like a public park or a library, meaning that we are accessible to everyone; we want to have knowledge that represents the full diversity and scope of the world in the languages that everyone speaks, in the experiences that people have. We want people to participate in knowledge – not just to access it, but to also be participants in the creation and curation of it. And we want to reach everyone. And those are things that we think of as critical to our vision.

We adhere to a strong sense of responsibility for accuracy. So one of the things about Wikipedia is, as the encyclopedia anyone can edit, we have been working very hard for the last 16 years to build a degree of trust and responsibility to our users to ensure that information on Wikipedia is accurate and reliable. And when we find information that is inaccurate, Wikipedians actually delight in that knowledge, because it means they can fix it. So, that’s quite critical to who we are.

I think one of the things that is very interesting today in terms of the overall knowledge environment is there is quite a discussion at the moment about what is accurate, reliable, trustworthy, neutral information. Wikimedia, over the course of the last 16 years, has worked very hard to be in a position to be able to answer some of those questions. The way that Wikipedia works is that articles that are included in Wikipedia have to be based on a set of core principles. They must be notable subjects, they must have been written about in secondary sources. Those secondary sources must be reliable and accurate. The content itself must be written in a neutral way, so that it is understood to have no particular bias or slant.

Of course, with 5 million articles on English Wikipedia alone, we are constantly in a state of self-improvement to live up to these policies.

As an encyclopedia that anyone can edit, the policies around reliability and neutrality and verifiability are quite important. You’ll see at the bottom of every Wikipedia article citations back to the sources from which that information comes. We think that that’s actually critical to why Wikipedia is trusted today, because if there is information on Wikipedia, you can actually see where did it come from; you can check almost every single edit that’s ever been made to any article – in fact, almost every single edit that’s ever been made to the entire website; you can understand why it was there, when it was placed there, who presented that information, where it comes from. And we believe that that transparency is critical to who we are. It’s core to our policies. It’s actually core to the platform itself, the technology that underlies the project. And that transparency creates a sense of accountability to our users.

So in this moment in time where we’re asking questions about where does information come from, who is creating it, why is it being presented, you can answer those questions on Wikipedia as you are actually learning about whatever it is you came there to learn.

We view ourselves as a layer of information on a much broader information ecosystem. Wikipedia is a general introduction to many different topics. We draw very heavily on the research and work of people who go out and discern information, advance research in the world, whether those are those in the academic sector, whether those are journalists, and we believe that what we are here to do is provide just an overview for people to learn more about the world around them.

And so that is just a little bit of an overview of who we are in the world, and I’d just like to open it up, I think, to any questions. I’d love to speak more about any of those points, but I’m actually here to answer your questions, whatever they may be.

MODERATOR: Thanks. If we can get you to identify yourselves on your first question. If any colleagues in New York are there, just jump in and make yourselves known, please. Thank you.

QUESTION: My name is Martha Andres. I am a correspondent for Prensa Latina news agency. I’m very interested in the part of the volunteers that you work with. I mean, how do you get to have the work done? How do you organize the work of the people there is --

MS MAHER: Certainly.

QUESTION: -- with that kind of research and anything?

MS MAHER: So there is an expression that we use at Wikipedia that I hope you find humorous: “It is a miracle that it works in practice because it would never work in theory.” The idea that you could organize hundreds of thousands of contributions in a given month and that the end result would be an encyclopedia is a fairly remarkable thing.

The way that we are organized is that Wikipedia is open to anyone to edit. People tend to organize themselves around topics that they’re interested in or roles that they would like to play. So, for example, if you’re interested in 18th century tapestries or military history or biology and the natural sciences, people will tend to gravitate towards improving articles around topics that they are interested in. And because it is inherently a social activity to edit Wikipedia, you get to know the other people who are contributing.

People will organize themselves around the work that needs to be done. Now, some of this happens in a very informal way, so I might see an error or a grammatical mistake and correct it as an individual. Some of this happens in a highly formalized way. Our medical community, which edits on medical-related topics, is actually so organized that they now have a foundation that they use to raise money to improve the quality of medical articles on Wikipedia. They have developed applications for mobile phones to be able to use medical Wikipedia offline in places where connectivity is poor.

So the way that it is organized is very much driven by the community itself, and people who pay attention to the needs of any particular area. Now, that all sounds quite open. We at the Wikimedia Foundation also support 100 global affiliates in many different parts of the world, ranging from Indonesia to Germany to Argentina to Ghana. You name it, there’s probably a Wikimedia near you. And those global affiliates are engaged in volunteer support. So they might organize events in which they focus on editing articles about specific topics, areas of interest, exchange of languages depending on where they are in the world. So it’s very much organized by the community itself.

What we view ourselves as here to do is to provide the resources for volunteers so that they can get the work done effectively.

QUESTION: Thank you.

MS MAHER: Of course.

QUESTION: I guess I have two questions. One is on maybe a more of a logistical side of it. So I see here more about the foundation, and it is a nonprofit. Maybe it says in here – who funds the foundation?

MS MAHER: Yes. So, we are funded by the public. We received, in 2015 - which is the latest numbers I have because we’re still closing out our 2016 year - we received donations from 5.4 million people from all over the world, every continent, and those donations are an average of about $14 per person, but they start at about $1 and go on up. So, part of our model is not just that we are created and written by anybody in the world, we are also supported by primarily small donors. So that fundraising accounts for 85 to 90 percent of our annual budget. The remaining 10 percent comes from larger donors, foundations, and institutions. We actually believe that that’s critical to maintaining our independence, is being supported by the same people that we serve.

QUESTION: Do you have a date when it became possible – in other words, at which point did the number of people contributing to Wikipedia made it possible to have sort of organized and hire staff and just --

MS MAHER: Ah, yes.

QUESTION: -- (inaudible) organization?

MS MAHER: So I believe – and I would actually want to check this – so the foundation – Wikimedia Foundation was created in 2003 when it became clear two years into the creation of the sites themselves that they needed some sort of structure to support them – to run the servers, for example. And the first fundraiser followed not far after that – believe it was within the first year of the foundation’s creation – and it was at the time to actually just purchase another set of servers because the projects and the sites were growing so quickly that it was difficult to support them given the conditions of their popularity.

Today, that – our budget has increased over time, as you can imagine, from a couple thousand dollars to run some servers to we are about $75 million a year in annual operating budget, which now supports both the running of the servers – as you can imagine, our traffic only continues to increase – but also the support for volunteer community. We have a grant-making program that goes to the volunteers who write the articles. We give away about 8 percent of our budget every single year in grants to the community and then provide other means of support – legal assistance and the like.

So, I guess the answer to your question is that we have been very lucky to see the funds scale around the same – at around the same rate as the popularity of the site itself.

QUESTION: Thank you. And the other question is more sort of specific, because some entities – sometimes countries, sometimes various other entities – are often engaged in conflict with each other.

MS MAHER: Mm-hmm, that is true.

QUESTION: And with the internet being what it is, all these people, very often their battles rage on the internet, including Wikipedia. So they enter information opposing each other and that’s an endless – an open-ended process --

MS MAHER: Debate, yes.

QUESTION: -- because they continuously present information from their standpoint.


QUESTION: So I did take note of what you said about ensuring accountability and like – and transparency and accuracy. So specifically these kinds of highly charged situations, what do you do to ensure some degree of accuracy and just to make sure that this – the process is not overtaken by people who are --

MS MAHER: Highly politicized points of view.

QUESTION: -- either highly politicized or highly passionate about a certain issue or otherwise are interested in turning – presenting the information their own way?

MS MAHER: It is a real challenge and it’s something that we face on a regular basis. One thing I will say is that you would be surprised at what becomes highly politicized, and I say this perhaps to inject some humor. I once received a request for comment from a reporter from – I think it was Le Monde – who had a question about a discussion that was happening on the English Wikipedia page for the city of Paris. What photo should illustrate the city of Paris? Should it be a photo of the historic core of the city or should it include a skyline with skyscrapers, because what is Paris’s identity? Is it a modern city or a historical city? And thousands of words had been written about this debate. So it is not just the most highly political issues of our day. Almost every part of knowledge has somebody as a stakeholder in it.

The way that this works – now to answer your question – the way that this works is that Wikipedians are – people identify as Wikipedians. They contribute to the projects over time. Those who are longstanding contributors are often recognized by their peers and given additional responsibilities on the site, administrative privileges. So as an ordinary contributor, you’re just a contributor. Then you can become an administrator, and the most senior level of this responsibility is to become a steward of the sites, and that is – there’s only about 30 of them in the world. They cover the world, they speak many different languages, and they have ultimate responsibility for paying attention to what’s happening across the sites.

With incidents such as the one – as any conflict that you might be describing, the way that Wikipedians tend to pay attention to these and resolve these issues is if there is a disagreement and that disagreement is unable to be quickly resolved, they may, for example, call what’s essentially a timeout, and they’ll pause editing on the page, and they’ll actually freeze it as the most neutral perspective that can be achieved. So a version of the article that is the most neutral and acceptable to all parties. And they’ll wait for people’s passions to cool down, and then they will reopen the page for editing.

Now, for some topics, passions never cool down. The discussion pages – behind every Wikipedia page, it’s helpful to think of it as a newsroom. There is what’s called a talk page. And the newsroom is where these discussions happen, and different people with different points of view can share their perspective.

What we have found is that over time, what generally happens is that people who contribute to the projects actually start taking on a more neutral point of view. There have been studies that show the longer you edit Wikipedia, even if you come in with a highly partisan perspective, the more likely you are to ultimately move towards a neutral - more neutral - perspective on issues.

An article can be locked for a day, it can be locked for an hour, it can be locked semi-permanently. That doesn’t mean that it isn’t possible to edit it; it just means that until you have been – have a certain level of a history within the community as a good-faith – what we would call a good-faith contributor, meaning that you have contributed with responsibility, you have contributed in line with the policies, you participated in what we call building the encyclopedia – you may still be able to contribute to that article. So a highly fraught issue may still be able to be edited by Wikipedians who have demonstrated a commitment to the encyclopedia and encyclopedic knowledge, but might actually be semi-locked or what you would call protected from editing from just random individuals, and that is actually one of the ways that we steward articles that are about particularly sensitive topics.

And of course, you can imagine – as whenever there’s an election anywhere in the world, the article of the pages of politicians, the articles around current events, breaking news – these tend to be articles that have a higher degree of scrutiny and protection.

Sorry, it’s a long and complicated, answer, but --

QUESTION: No, no, no, thank you. (Inaudible.)

MS MAHER: -- as you can imagine.

QUESTION: I don’t want to hijack the conversation, but this is the last question. It is sort of tied to what you just said. Because, again, with highly charged topics, there are two sides to every issue.

MS MAHER: Sometimes more than two.

QUESTION: Or sometimes more, and always somebody’s freedom fighter is somebody’s (inaudible) militia.

MS MAHER: Certainly.

QUESTION: And so do you – does Wikipedia on its own take a stand on these issues, or you just let the issue sort of – yeah.

MS MAHER: Yes. That’s a great question. One additional thing before I answer your question I should have mentioned at the outset is that often another way that Wikipedians will handle controversial issues is they will describe them in the article. So there may be – and I often use the example of climate change. The articles – there are multiple articles on climate change, but the majority of them describe the science, climate science, and then at the bottom of the article, they’ll have a section on controversy. And they’ll talk about what the controversies might be from a political standpoint, they’ll talk about discussions within the scientific community, areas of emerging understanding. And that’s very often how controversial topics are handled, is there’ll be a description of the known information and the historical information, and then a section that handles controversy.

To your question, the Wikimedia Foundation does not take an editorial line for the Wikimedia projects. We are not involved in the editorial policy of the projects at all. The policies that I described earlier around neutrality, verifiability, reference to reliable sources, have actually been determined by the editors themselves over time, and have been developed in response to ensuring that the information is as accurate and high-quality as it can be.

Now, I say that; we are always in the process of self-improvement. You will find errors on Wikipedia. It is why we encourage you to check the citations and follow up and do additional research. We are just meant to be an overview.

But the Wikimedia Foundation is not involved in setting editorial policy. We have found that the community is quite responsible and takes into account some of these issues and has found ways to handle it over time.

QUESTION: Thank you.

MS MAHER: Of course. Thank you.

QUESTION: How do you deal with the complaints? I mean, I suppose you receive a lot of complaints or a lot of critiques from the people that are reading a certain article and do not agree with what is said there. How do you deal with that kind of situation?

MS MAHER: Absolutely. When we receive complaints or concerns, we generally encourage people to engage in the article itself. If they see something that is wrong, please, we would ask them to edit it, and edit it with a citation to a reliable source, something that has been peer-reviewed, something that has been published, gone through an editorial process. The way that we think about reliable sources is very much the characteristic of how those sources create, generate, and hold themselves accountable to the knowledge that they produce.

So we would encourage somebody, if they see something wrong, please fix it. It is the encyclopedia anyone can edit.

Now, when it comes to issues of, say, a biography of an individual who – and that individual might not like what’s written on the biography, Wikipedia does have some very strong policies in place around protections of biographies of living individuals. We take that very seriously because obviously, as one of the world’s most visited sites, what is written on a Wikipedia page actually becomes very much what people know about you. So there are some very detailed policies about what should and should not be included in a biography of a living person. If there’s something that is inaccurate, we encourage people to engage on the talk page to make the editors aware of that inaccuracy and ideally to provide additional information.

But generally speaking, Wikipedians are much more responsive to concerns about inaccuracy. There – it is true that there are unpopular facts in the world, and there are things that people perhaps don’t wish were true. That doesn’t mean they don’t belong in an encyclopedia. In fact, very often that’s exactly where they belong.

QUESTION: I’m Jim Lobe from Inter Press Service. You said there have – when an entry is put under controversy – that is, there is a note on the top --

MS MAHER: Mm-hmm.

QUESTION: -- who decides whether it has that (inaudible)?

MS MAHER: Yes, the tag at the top that says “this article is disputed or needs additional citations.”


MS MAHER: Those tend to be those editors that I mentioned who have more experience and time within the projects. And so as much as I – there might be people who want to focus, say, on popular culture and edit articles about television shows, there are also editors who just focus on the quality of the articles themselves and ensuring that they adhere to certain standards. And so that’s actually driven by the editor – by the editorial community. That’s not – the Wikimedia Foundation does not get involved in tagging the articles for their accuracy. That’s very much driven by the community of creators.

QUESTION: So you’ve indicated there was some kind of hierarchy with respect to longstanding contributors who become administrators, and then there are 30 --

MS MAHER: Stewards.

QUESTION: It sounds like a high council.

MS MAHER: (Laughter.) I should be clear that it’s not – they actually are not involved in the content creation of stewardship. They do things like protect the projects from spam and ensure that if there are issues with regards to the way that the sites are being governed, that they’re available to step in and provide additional resources as necessary. Content itself is very much driven by the editorial community, and that includes article tags that say things like “this article – the neutrality of this article is disputed” or “this article needs additional citations.” That is – as an editor, you could actually add those tags if you so chose.

QUESTION: And how does one contribute, like, to the talk page, or how does one introduce oneself?

MS MAHER: So you can create an account, or you could not create an account – both of those are possible. You don’t need to create an account to edit Wikipedia. The top of every article has two tabs on the left-hand side and a number of other tabs on the right-hand side. The two tabs on the left-hand side are the article page and the talk page, and so you can check the talk page of any article as you are looking at it. If you would like to edit it, you go over to the right-hand side and click the edit button, and that is true both of the article and of the talk page, and that’s it. You then can write whatever it is you’d like to write.

QUESTION: And with respect to neutrality, especially given an increasingly partisan --

MS MAHER: Conversation in the public, yep.

QUESTION: In this country, anyway.

MS MAHER: Not just in this country, certainly.

QUESTION: Like, if a word like “conservative” is used, does there have to be – is there some understanding of what that word means? I cite that just as an example where --

MS MAHER: Certainly.

QUESTION: -- Steve Bannon is described as a conservative, which strikes me as totally bizarre.

MS MAHER: I haven’t --

QUESTION: No, no, I don’t know that --

MS MAHER: Oh, no, I’m just saying I’m curious. I don’t know what his Wikipedia page would say. What we would generally – Wikipedians, the approach that one might take is to remove adjectives that are in dispute or overly evocative of a single perspective. Because the goal of neutrality is to focus on information that’s known, facts that are proven and in evidence, and descriptors – including “conservative” or “liberal,” which I think are general parts of the spectrum no matter where you are in the world – might be seen as not neutral in the description of an individual’s personal political identity. Of course, if it was used to describe, say, a policy platform for a political party, that might be appropriate. And so Wikipedians will use the contextual information and tend to remove information that is too adjectival in nature because it is not necessarily neutral in the way that it’s described. And I edit Wikipedia; I’ve done that myself – gone in and said, ‘gosh, this doesn’t sound very neutral, let me take out these qualifying adjectives and just stick to the information that’s available and evidence and cite it back to reliable sources.’

QUESTION: And also with respect to your more longstanding and hence more influential contributors, I mean, is there any kind of demographic profile that you’re trying to take? I mean, arguably, people who have more time to edit may be somewhat better off just as like --

MS MAHER: You are absolutely correct.

QUESTION: -- in the glorious ‘50s, women contributed much more to civic society because they were generally homemakers and --

MS MAHER: Certainly.

QUESTION: -- I mean, which was --

MS MAHER: No, no, these are very interesting questions. So I think that there – it’s important to note that given that we exist in 300 languages and have people and Wikipedians who contribute from all over the world – I mean, from the hills of Nepal to down here sitting in D.C. to McMurdo Station in Antarctica – it is truly everywhere. It’s hard to characterize the demographics of any one Wikipedian. But broadly speaking, we would note that Wikipedians tend to have more leisure time, as you would note, so we can infer that they are people of higher socioeconomic status in whatever society they live in. That doesn’t necessarily indicate a gender, it doesn’t necessarily indicate a class per se, but we would acknowledge that a degree of leisure time in order to participate in knowledge creation is likely an indicator of higher socioeconomic status. What this means, and we are very open about this, is that we would like to see more people of greater diversity of backgrounds participating on Wikipedia.

We know, generally speaking, more males than females participate. We view this as an imbalance, and we view the resulting bias in terms of what is covered on Wikipedia as an outcome of this. So an example is that of the 1.3 million articles on English Wikipedia that are biographies, only 16 percent of those are about women.

Clearly, that does not reflect the balance of our demographics in society no matter where you are. That is likely for a variety of reasons. One, Wikipedia is a mirror held up to the world. These sources that we have available to write articles about individuals, about things in the world, are very much determined by what people have written about, the research that’s been done, what journalists cover. If the information doesn’t exist in a reliable secondary source, it’s very hard for it to be on Wikipedia. So there is an information bias that exists. We reflect that. We reflect the biases of society.

At the same time, we know that a more diverse pool of editors who come from different backgrounds, have different identities, different life experiences leads to a richer and more inclusive understanding of what knowledge is and what matters and what should be included in the encyclopedia.

So one thing the Wikimedia Foundation is very focused on doing is increasing the diversity of our editing community however that is defined, because the diversity of the editing community in the United States may be a very different thing than the diversity of the editing community in India or in South Africa. And so we’re always looking at how we think about diversity within the context of a language group, within the context of multiple language groups, within the context of a national background or a regional identity.

QUESTION: And then what steps are you taking to try to enhance that (inaudible)?

MS MAHER: Certainly. So the Wikimedia Foundation, as the foundation I mentioned, we have our operating budget. About 50 percent of our operating budget goes to supporting the sites themselves, making sure that they’re easy to edit, making sure they’re available, reliable, secure to anyone in the world, [and] protect the privacy of the individuals using it. The other portion of what we do is we run programs that support the editing community, and some of that includes, as I mentioned, grants. We have allocated about a quarter of a million dollars every year specifically to increasing diversity and inclusion within the overall Wikipedia community. We have taken steps that support – we have 100 global affiliates around the world. Those global affiliates often have programs that include outreach to communities that are underrepresented. That includes everything from understanding oral histories in certain places where that’s appropriate to running edit – what we call edit-a-thons, around increasing the participation of women in Wikipedia or the representation thereof.

So we take various different steps as the foundation to support diversity. We are also very much focused on the dynamics of contribution on the projects, making sure that it’s a healthy place to contribute, making sure there is not a – that we’re reducing harassment so that it feels as though everyone is welcome. These are additional things that we do. But I want to also call out how much this has very much been driven by the contributor community themselves. Once the contributor community was aware of some of these systemic biases that exist, they – there have been a number of initiatives from within the editing community to actually take these on. So self-organized events, whether it is around improving the quality of women, as I mentioned; improving the quality of African – coverage of African-American history; improving the quality of coverage of underrepresented castes, it really is very much highly determined by the editing community themselves.

QUESTION: In your opinion, what – which are the highest risk do you take by working with this amount of information every day? I mean, this – where one have so much information than – on the internet, where people are pushing all the time different things and there is this discussion about what is accurate and what is not accurate. What are the highest risks that Wikipedia faces?

MS MAHER: I think the most important thing is to remain a place that people trust and to focus on how to sort through fact and fiction in order to be able to present a single, neutral perspective on information, even when it is contentious and even when it is difficult. I have a lot of confidence in our ability to do that, because we’ve been doing it for 16 years now. The issue of fake news or the issue of misinformation is actually not new in the world; it’s just particularly important and influential in part – in our sort of discourse globally at the moment. But it has always existed. Bad information has always existed. Misinformation has always existed. This is – ever since the printing press was created, ever since the first press freedom laws in Sweden 250 years ago, there have been issues around is information accurate, is it reliable.

I think that Wikimedia has developed some great policies over time in order to address this. I think also one of the strengths of the model is that it can continue to evolve. It is very much built in relation to the world around it. But I do think that it is a priority and I think it’s something that our editing community is very much focused on, is ensuring that neutrality – ensuring that it’s a trusted resource for people globally.

On an individual level, working with information can at times pose a challenge to the individuals who create it, and so safeguarding the privacy and security and identity and the safety of the individuals who contribute to Wikipedia in general I think is also critical in order to ensure that information can continue to be presented, even when that information perhaps is unpopular.

QUESTION: To just follow up on something you said about ensuring diversity and understanding that some people, just by virtue of having a little more time or (inaudible) reason, may be contributing more. And in some situations when there is an ethnic conflict, say, between the two sides, and historically or internationally one of the two warring sides happens to have a more present English-speaking community versus the other that doesn’t, that obviously affects how much information is put out there even though Wikipedia exists in all languages. But if you take the English language sphere, in the – within the same articles, there will be more editing by one side just – which just has more English speakers or communities, say, within the United States (inaudible). So is that ever taken into account in terms of ensuring diversity or trying to ensure that the issues --

MS MAHER: That’s a great question. I think that English Wikipedia is a very unique case as one of the languages that is so widely spoken globally, with contributors from Singapore to Canada to the UK to South Africa. It is an unusual example because of the fact that it can be contributed to by so many different people. We are aware that sometimes this is a concern. I think that one of the things that usually happens in that instance is that there is an emphasis within the long-term contributor community, which is often different than people who are there solely for one political perspective, to, as I mentioned earlier, take steps to protect articles, to prevent them from being overly biased in one way or another. Very often Wikipedians will actually make an effort to bring somebody in who knows nothing about the topic and so therefore is not – does not have a particular perspective on it in order to adjudicate those issues and come to some sort of resolution and perhaps, as I said, put a pause on the conversation.

We are aware that this is a challenge. We are aware that information is not distributed equally, that access to the internet is not distributed equally, that access to English-speaking competence or fluency is not distributed equally. It’s something that I think is part of the process of continual reinvention and awareness. And the more that we are aware of it, the more likely we are to be able to address it. I always say that Wikipedia’s only in competition with itself – in competition with itself to become better and in competition with itself to be more accurate, in competition with itself to be more neutral. It is an ongoing process, and we’re very much aware and sort of acknowledge that.

QUESTION: Have you banned people from publishing things in Wikipedia, maybe people you know that have usually who make some kind of mistake or have not great information, then do you ban people from doing the --

MS MAHER: The Wikimedia Foundation does not, generally speaking, get involved in editorial decisions, such as bans and blocks. I will say that about two and a half years ago a Wikipedian who had administrative privileges, which as I mentioned are conferred by the community itself, so it’s through a process of consultation – so anyone with administrative privileges has them because they are trusted by other members of the community – actually blocked the entire House of Congress of the United States from contributing because they were vandalizing Wikipedia and writing information that was inaccurate and not constructive to the encyclopedia. And so yes, people get blocked all the time. And sometimes institutions you would not expect get blocked.

The blocks are very much focused on constructive contribution to the encyclopedia. If you are constructively contributing, you will not be blocked. If you are putting in misinformation or bad information or simply vandalizing it because your sports team won and you want to acknowledge how great they are, which is one of the most common forms of vandalism, you run the risk of being blocked, yes. And those blocks may be short. They might be an hour; they might be a day. If you’re particularly egregious as a vandal, it might be a little bit longer.

QUESTION: And what about when people contribute pseudonymously, as they use different names in order to (inaudible)?

MS MAHER: Most people do contribute pseudonymously. Very few people use their real names. That’s okay. The interesting thing about pseudonymity is that it still has identities attached to it. You can have a sense of what somebody is, and that is often how the editing community works. We allow people to edit anonymously, which leaves the record of their IP address, so it is anonymous, but that is a decision to allow that data to exist out there in the public, or to create an account – and which can be a pseudonym, or to use their real name. Those are all options.

When people edit pseudonymously and may have more than one account, that is frowned upon. That is called sock-puppeting, in the vernacular. And if one is caught doing that, one can be blocked.

QUESTION: With respect to that specific article or like – there are farther (inaudible)?

MS MAHER: There are escalating levels of sanctions, depending on how bad faith an actor you are. Wikipedians refer to good faith or bad faith. Good faith is building an encyclopedia. Bad faith is causing harm.

QUESTION: And then how – is that also determined by – simply by edit spontaneously through the editorial community, or is that – does that go to another level?

MS MAHER: It can be appealed. Again, it sort of depends on how egregious an issue is. We have found efforts – sock-puppeting rings – to try to influence information. Very often the biggest issue for us in this case is public relations firms working on behalf of commercial clients who are trying to improve their image on the site. And that is a violation of our terms of use, and so that can also lead to a block or a ban.

QUESTION: And so how do you determine that then?

MS MAHER: There are editors who spend a lot of time researching this, and that is what they work on almost exclusively.

QUESTION: And so they work across --

MS MAHER: They’ll work across the projects and track – as it turns out, when you’re – this is not a new problem on the internet. There are actually sort of signatures that you leave behind when you’re engaged in this sort of behavior. They tend to be quite common in nature. And so the patterns are highly replicable, and so people will sit – engage in protecting – protecting the Wikimedia projects as volunteers and actually track this sort of editing behavior in order to prevent it from happening. It’s pretty remarkable, if you think about it.

QUESTION: I’m getting --

MS MAHER: I’m sorry. I’m just personally in awe of the fact that people do this. It’s really quite remarkable.

QUESTION: No, I mean, I don’t want to take too much time, but do – so there are people who look over the entire network and they look for certain patterns. They’re not brought into the – a conversation, say, the top page --

MS MAHER: They might be. If there’s an issue and people suspect that somebody is engaged in sock-puppeting, as we call it, yes, people might be brought in to investigate and try to understand whether that is – in case – whether that is, in fact, happening or not. There are people who specialize in it, sort of detectives of Wikipedia.

QUESTION: I have two other questions.

MS MAHER: Please.

QUESTION: This is very interesting because – and a tool we use – we use it so often and sometimes we want to know what has happened and we have that information.

MS MAHER: Yes. Most people just see the articles and have no idea of the complexity behind it.

QUESTION: Yes. One is related to how is the process of translating? And I know you have continued this --


QUESTION: -- but I ask you when someone, for instance, will be something in Spanish, do you have something that translate that into English or French --

MS MAHER: That’s a great question.

QUESTION: -- or other language? How does it work – that process work? And how do you deal with the publishing of hate comments, for instance, or that kind of information that is not only bad for the – for Wikipedia, but also even it may have legal issues related to it?

MS MAHER: Some great questions. So I’m going to answer the first, and then I’ll try to answer the second. If I forget, please remind me. So the – actually, I’ve forgotten the first. I’m sorry. So the first question – oh, language translation. So there is no direct translation. The English Wikipedia, which is the largest, has around 5.4 million articles, and the next largest Wikipedias are, I think, around 2 million articles, and then there’s a number that are around that size, 1 to 2, and then they get smaller from there.

We have tools that allow editors to translate from language to language, and it is not necessarily from English to Spanish. It could be from Azeri to Spanish. It could be from Persian to Chinese. It really – it depends on what people want to use. Those tools allow articles to be translated from one language to another. We do use machine translation to facilitate that, although it’s not necessary or required.

Different languages take different approaches. The Swedish Wikipedia, for example, is a very large Wikipedia because it does engage with the replication from other Wikipedias, but that is a choice the Swedish community has made. The English community and the German community don’t do that. That’s a policy that they have decided on.

So, it’s highly idiosyncratic, depending on the language community. Some smaller languages have taken the approach of wanting to grow faster. Vietnamese is one of these languages. And so it has actually used a series of bots to replicate what we would call stub articles, so just the title and perhaps a single sentence, which can then be improved upon later. Others – I use German as the example. They are deliberately intended to be a smaller encyclopedia. They’re more focused on quality than quantity.

It is possible, of course, to think about translating across all these different languages. We are very interested in what is actually contained across the whole ecosystem of 300 languages, because there is no one superset. English Wikipedia is not necessarily the most comprehensive; it is just the Wikipedia that is – reflects the interest of an English-speaking community. There is information in Swahili Wikipedia that is certainly not in English Wikipedia, and vice versa. So that’s the answer on language translation.

Now, the second question was around hate speech or hate comments. Those would not generally last in an article for very long. There are ways – the way that the encyclopedia works is that individual editors are focused on a topic. I like to use the example of the article page in English for HIV and AIDS. There are more than a thousand people who receive an alert every single time that article is edited, so more than a thousand people get a notification when a change is made and go check to see if that is a good change or a bad change. So very quickly, changes that are vandalism, changes that are inaccurate, changes that are focused on information that doesn’t belong in an article itself, are very quickly reverted back to the original state of the article.

The issue of conversations that happen in the talk pages is something that we are actually focused on, mostly around creating a more pleasant environment for people to contribute. And so we recently received a gift from the Craig Newmark Foundation to be able to improve the tools that exist for editors to track conversations that are hostile in tone and address those when those arise. So it is something that we are focused on. Right now it is very much a manual process, and we’re looking at ways to actually support contributors so that it is less manual, because it takes quite a bit of time. But generally speaking, when those sort of things exist, they are very quickly deleted. It’s against our policies to have – to engage in uncivil speech.

MODERATOR: Shall we take a final question, and then we’ll allow any personal interactions?

MS MAHER: Sorry, I know it’s a lot of information.

QUESTION: I wanted to clarify on the language issue. Because I work with Russian all the time; I work for a Russian-language publication. I obviously look for information in English and Russian. And the Wikipedia entry for a certain public figure --

MS MAHER: Will be very different.

QUESTION: -- will be very different in Russian and in English. So there’s basically – unless somebody from the community puts those – makes them sort of similar or equivalent, there’s nobody at the Wikipedia itself that will --

MS MAHER: That’s correct. There’s nobody --

QUESTION: -- make any changes?

MS MAHER: No. The Wikimedia Foundation, as I said, we don’t actually edit the content at all. We’re here to support the sites, make sure they run, make sure they’re fast, make sure they’re accessible, make sure that the people who edit the sites have the time and resources that they need in order to contribute. Some community members are interested in doing language-to-language consistency and translation, but the reality of the matter is that they are going to be different. The interests of different language communities are different, and we actually see that as a strength rather than a weakness. It does mean that on certain topics where there are differing political perspectives, you may see those biases emerge.

I’m very interested, as technology continues to improve and advance, the ability of machine translation to perhaps provide us more insight into that. I think that that will be part of our continued evolution is trying to understand what exists in one language versus another. I think that is a very interesting journey for us to take. As I said earlier, we’re constantly in the process of improving, and that might be an interesting area for our community to focus on.

MODERATOR: So with that, we’ll conclude the briefing. We want to thank our briefer, Katherine Maher, for coming --

MS MAHER: Thank you.

MODERATOR: -- all the way here to join us, and we appreciate the small group of you that have come to attend. I think that made for a more robust discussion. Thank you very much for joining us.

MS MAHER: Thank you very much for having me and for all your good questions.