The Case of Philip Roth vs. Wikipedia

As Wikipedia becomes an increasingly dominant part of our digital media diet, what was once anomalous has become a regular occurrence.

Someone surfing the net comes face to face with a Wikipedia article — about himself. Or about her own work.

There’s erroneous information that needs to be fixed, but Wikipedia’s 10-year-old tangle of editing policies stands in the way, and its boisterous editing community can be fearsome.

If a person can put the error into the public spotlight, then publicly shaming Wikipedia’s volunteers into action can do the trick. But not without some pain.

The most recent episode?

The case of Pulitzer Prize winning fiction writer Philip Roth.

His bestselling novel “The Human Stain” tells the story of fictional character Coleman Silk, an African-American professor who presents himself as having a Jewish background and the trials he faces after leaving his university job in disgrace. Widely read and highly acclaimed, the book was reviewed or referenced by many famous writers, such as Michiko Kakutani and Janet Maslin of the New York Times and the noted Harvard professor Henry Louis Gates, Jr. [1] [2] [3]

The Broyard Theory

But there was a standing mystery about the novel.

After the book’s release in 2000, Roth had not elaborated on the inspiration for the professor Silk character . Over the years, it had become the subject of speculation, with most of the literary world pointing to Anatole Broyard, a famous writer and NY Times critic who “passed” in white circles without explicitly acknowledging his African American roots.

In 2000,’s Charles Taylor wrote about Roth’s new book:

The thrill of gossip become literature hovers over “The Human Stain”: There’s no way Roth could have tackled this subject without thinking of Anatole Broyard, the late literary critic who passed as white for many years.

Brent Staples’ 2003 piece in The New York Times wrote that the story of Silk as a “character who jettisons his black family to live as white was strongly reminiscent of Mr. Broyard.”

Janet Maslin wrote the book was “seemingly prompted by the Broyard story.”

It was such a widely held notion, the Broyard connection was incorporated into the Wikipedia article on “The Human Stain.”

An early 2005 version of the Wikipedia entry cited Henry Louis Gates Jr., and by March 2008, it relayed the theory from Charles Taylor’s review.

The view was so pervasive, a list of over a dozen notable citations from prominent writers and publications were found by Wikipedia editors.

Wikipedians researching the topic came across articles as secondary sources that drew parallels between Silk and Anatole Broyard. The references were verifiable, linkable prose from notable writers and respected publications. The core policies of Wikipedia — verifiability, using reliable sources and not undertaking original research — were upheld by using reputable content as the basis for the conclusions.

Roth Explains It All

However, information from Roth in 2008 changed things.

Bloomberg News did an interview with the author about his new book at the time, “Indignation.” Towards the end of the interview, he was asked a casual question about “The Human Stain:”

Hilferty: Is Coleman Silk, the black man who willfully passes as white in “The Human Stain,” based on anyone you knew?

Roth: No. There was much talk at the time that he was based on a journalist and writer named Anatole Broyard. I knew Anatole slightly, and I didn’t know he was black. Eventually there was a New Yorker article describing Anatole’s life written months and months after I had begun my book. So, no connection.

It might have been the first time Roth went on the record saying there was no connection between the fictional Silk and real-life writer Broyard. It seems to be the earliest record on the Internet of this fact.

Fast forward to 2012, and according to Roth, he read the Wikipedia article for [[The Human Stain]] for the first time, and found the erroneous assertions about Anatole Broyard as a template for his main character. In August 2012, Roth’s biographer, Blake Bailey, became an interlocutor who tried to change the Wikipedia entry to remove the false information. It became an unexpected tussle with Wikipedia’s volunteer editors.

Unfortunately for Roth, by the rules of Wikipedia, first-hand information from the mouth of the author does not immediately change Wikipedia. The policies of verifiability and forbidding original research prevent a direct email or a phone call to Wikpedia’s governing foundation or its volunteers from being the final word.

Enter The New Yorker

Frustrated with the process, Roth wrote a long article for the New Yorker, detailing his Wikipedia conundrum. He provided an exhaustive description of the actual inspiration for the professor Silk character: his friend and Princeton professor, Melvin Tumin.

“The Human Stain” was inspired, rather, by an unhappy event in the life of my late friend Melvin Tumin, professor of sociology at Princeton for some thirty years.

And it is this that inspired me to write “The Human Stain”: not something that may or may not have happened in the Manhattan life of the cosmopolitan literary figure Anatole Broyard but what actually did happen in the life of Professor Melvin Tumin, sixty miles south of Manhattan in the college town of Princeton, New Jersey, where I had met Mel, his wife, Sylvia, and his two sons when I was Princeton’s writer-in-residence in the early nineteen-sixties.

Good enough. But the problem arose when Roth attempted to correct the information in Wikipedia with the help of Bailey, his biographer. He wrote:

Yet when, through an official interlocutor, I recently petitioned Wikipedia to delete this misstatement, along with two others, my interlocutor was told by the “English Wikipedia Administrator”—in a letter dated August 25th and addressed to my interlocutor—that I, Roth, was not a credible source: “I understand your point that the author is the greatest authority on their own work,” writes the Wikipedia Administrator—“but we require secondary sources.”

Thus was created the occasion for this open letter. After failing to get a change made through the usual channels, I don’t know how else to proceed.

The frustration is understandable. That someone’s first-hand knowledge about their own work could be rejected in this manner seems inane. But it’s a fundamental working process of Wikipedia, which depends on reliable (secondary) sources to vet and vouch for the information.

Because of this, Wikipedia is fundamentally a curated tertiary source — when it works, it’s a researched and verified work that points to references both original and secondary, but mostly the latter.

It’s garbage in, garbage out. It’s only as good as the verifiable sources and references it can link to.

But it is also this policy that infuriates many Wikipedia outsiders.

During the debate over Roth’s edits, one Wikipedia administrator (an experienced editor in the volunteer community) cited Wikipedia’s famous refrain:

Verifiability, not truth, is the burden.
– ChrisGualtieri (talk) 15:53, 8 September 2012 (UTC)

By design, Wikipedia’s community couldn’t use an email from an original source as the final word. Wikipedia depends on information from a reliable source in a tangible form, and the verification it provides.

Reliable sources perform the gatekeeping function familiar in academic publishing, where peer review guarantees a level of rigor and fact checking from those with established track records.

But even with rigorous references, verifiability can be hard.

Consider Roth’s New Yorker piece, where he says:

“The Human Stain” was inspired, rather, by an unhappy event in the life of my late friend Melvin Tumin, professor of sociology at Princeton for some thirty years.

Compare that to the 2008 interview, when asked, “Is Coleman Silk, the black man who willfully passes as white in “The Human Stain,” based on anyone you knew?” Roth said, “No.

This would seem to contradict the New Yorker article. This doesn’t make Roth dishonest. Rather, Roth likely interpreted the question differently in a spoken interview as to whether he knew anyone who “passed” in real life, as Silk did in the novel.

The point of all this?

Truth via verification is not easy or obvious.

Even with multiple reliable sources — a direct transcript from an interview or the words from the author himself — ferreting out the truth requires standards and deliberation.

As of this writing, Roth’s explanation about the Coleman Silk character has become the dominant one in the Wikipedia article, as it should be.

However, the erroneous speculation about Anatole Broyard was so prevalent and widely held in the years before Roth’s clarification, that it still has a significant mention in the article for historical purposes. There’s still debate how prominent this should be in the entry, given that it’s been flatly denied by Roth.


Roth’s New Yorker article caused the article to be fixed, but getting such a prominent soapbox is not a solution that scales for everyone who has a problem with Wikipedia.

After a decade of Wikipedia’s existence as the chaotic encyclopedia that “anyone can edit,” its ironic that its stringent standards for verifiability and moving slowly and deliberately with information now make those qualities a target for criticism.

Wikipedia has been portrayed as being too loose (“Anyone can edit Wikipedia? How can I trust it?”) and too strict (“Wikipedia doesn’t consider Roth a credible source about himself? How can I trust it?”). The fact is, on balance, this yin-yang relationship serves Wikipedia well the vast majority of the time by being responsive and thorough — by being quick by nature, yet slow by design.

It continues to be one of the most visited web properties in the world (fifth according to ComScore), by refining its policies to observe the reputation of living persons and to enforce accuracy in fast-changing articles. Most outsiders would be surprised to see how conscientious and pedantic Wikipedia’s editors are to get things right, despite a mercurial volunteer community in need of a decorum upgrade and the occasional standoff with award-winning novelists.

Andrew Lih is an associate professor at the University of Southern California’s Annenberg School of Communication and Journalism where he directs the new media program. He is the author of The Wikipedia Revolution: How a bunch of nobodies created the world’s greatest encyclopedia, (Hyperion 2009, Aurum UK 2009) and is a noted expert on online collaboration and participatory journalism. This story also appeared on his personal blog.

5 lessons learned: Improving civic engagement through a local news site

Four years ago a team of communication scholars, researchers and journalists set out to create a community news website that would increase civic engagement and cross ethnic barriers in a predominantly Asian and Latino immigrant city. Since Alhambra Source launched in 2010, it has grown to more than 60 community contributors who speak 10 languages and range in age from high school students to retirees. Their stories have helped shape local policy and contributed to a more engaged citizenry within a diverse community. Below are five lessons we’ve learned about creating a community news website that fosters civic engagement.

1. Investigate your community’s news and information needs before you launch.
While few news organizations are likely to have a dedicated team of researchers and scholars at their disposal, they can — and should — identify community information needs to guide the development of their site. On the simplest level, that means a reporter should know his or her beat well and do some investigating before launch.

As a journalist in Alhambra, for example, I witnessed firsthand the civic participation gaps and the barriers between ethnic and linguistic groups that our researchers had identified. The lack of civic participation was made evident in 2010 when five incumbents ran unchallenged, prompting officials to cancel the elections.

The need to cross language lines became clear when school and government officials, police officers and other community leaders all told me that they could not understand the most active press coverage of Alhambra: the Chinese-language newspapers. These newspapers target about a third of the city’s population, and yet city leaders had no idea what was being reported. Identifying basic communication needs such as these can help define the goals of a local news source and also establish a baseline that can later be used to demonstrate the site’s impact to funders or other supporters.

2. To effectively build a community contributor team, hold regular meetings, play to contributor strengths, and remember they are volunteers.
We work with community contributors — in our case that means Alhambra residents who volunteer and tend not to have professional journalism experience. Initially, I set about recruiting Alhambrans to report stories that might interest them or their neighbors. I searched for people already producing content online, talked to leaders of community organizations, and spread the word about our new site. Once we launched the site, we featured our contributors prominently with a call for others to get involved.

Monthly meetings in our office space have been crucial to the strength and expansion of our team. They are part newsroom story meeting, part community advocacy, and part social gathering (we always include a potluck dinner). After the first few meetings and the site launch, I no longer had to actively recruit contributors — at least one new candidate would contact me each month. As our reputation grows, so has our team. That doesn’t mean everyone sticks around: like any volunteer community, we have to work to keep people engaged and interested in giving their time. But enough new people come to keep up the site’s content and energy, while a regular base of contributors provide a core continuum.

3. When it comes to community contributions, a personal perspective is often crucial to a story.
Community contributors often want to report because they have an agenda they want heard. Obscuring that under a veil of objectivity just does not work on a community level. I’ve found community contributors are great for insight stories and features, sometimes providing our most creative articles, ranging from a critique of the local food rating system (“A=American, B=Better, C=Chinese”) to a call for new bike laws to a visit to the local psychic “Mrs. Lin.”

One story type that I have found community contributors can consistently produce better than outside reporters is a first-person piece incorporating a wider perspective. The stories that have received some of the highest traffic on our site and met our research metrics of increased civic engagement have tended to be of this type. Some examples include a story on the challenges of inter-generational communication for a child of immigrants, one about growing up Arab or Muslim in a mostly Asian and Latino community, and one about why a church community organizer takes issue with a city ordinance.

Finally—and this is important—keep in mind that these are not professional reporters. Everyone needs an editor, and working with community contributors often means multiple drafts and intensive fact checking. Many times it would have been easier for me to have done the story myself, so it is important to match volunteer reporters with pieces to which they can add value.

4. Crossing language and ethnic divides cannot be achieved through multilingual content alone.
Before we launched, we intended to be a site in the three languages most spoken by our readers — English, Chinese, and Spanish. We quickly discovered that we lacked the resources. And as it turns out, such a plan might not have been worth the effort.

About a quarter of Alhambra residents live in households where no adults speak fluent English. There is a clear need for foreign language media, particularly in the ethnic Chinese community. But that does not mean that the community would be interested if we created a multilingual website. From anecdotal interviewing, we found that these residents are satisfied getting their news from ethnic publications and are less likely to go to a website.

Instead, we found many other important ways to bridge the language divide. Here are four:

  • Building a multilingual team, which helps expand the range of stories we can cover and the types of people we can interview
  • Translating local foreign-language coverage into English
  • Translating selections of our own original content into Spanish and Chinese (through two means: high-quality human translations for select articles and Google Translate function across the entire site)
  • Establishing relationships with ethnic press so they print versions of our articles in their newspapers.

5. Use feedback loops as engagement and learning tools.
We use polls and surveys extensively on the site to engage residents, create a link between them and city officials, and improve our coverage. Some of our most successful surveys have ranged from where to find the best local burger or boba to whether the city should ban fireworks sales to which supermarket should come to Main Street.

We often incorporate the findings from these informal polls into stories. It enables more residents to participate on the site in a simpler way than writing a story, and in public policy issues, it offers a means for us to share community feedback with the government. For example, when the city council recently acted to limit pay-for-recycling, less than a handful of people from the public came to the meeting (like most days). But on our site more than 100 people voted to express their opinions, the vast majority against the ban. The city council then decided to grant a reprieve to one market.

We also use the polls to gauge our impact and to see on which topics residents would like more coverage. We have surveyed residents about what stories they would like to see, research questions they would like answered, and even improvements we could make to our website. Engaging the community this way enables us to better respond to their needs. After all, a community news site, like a city itself, is a work in progress.

Alhambra Source is the pilot project of a new Civic Engagement and Journalism Initiative at USC Annenberg. The project aims to link Communication research and Journalism to engage diverse, under-served Los Angeles communities. The Metamorphosis Project is the primary researcher, and Intersections South LA is another project site. This is the first in a series of articles on the topic of creating and evaluating local news websites that strive to increase civic engagement.

National party conventions, graphic photos, social media's bull$#!t, open data, and a world stream

Here’s a quick roundup of stories and conversations that caught our attention in the past week, the first in what will gradually become a regular series.

Convention City: For the next two weeks, we’ll be barraged with reportage from the Republican and Democratic national conventions. As MediaShift points out, a lot of attention among media observers will be paid to how a variety of digital tools are deployed, much like it was during the Summer Olympics. The media industry blog has already put together a helpful list of resources for following the conventions. Meanwhile, the Washington Post has launched a new feature it’s calling The Grid, which is an interesting way to scan through all their various social media and reporting channels and get the latest on the RNC (and next week the DNC).

Instagraphic: In case you missed it (which seems impossible), Instagram moved to the center of a century-old debate this weekend following the shootings at the Empire State Building. When user @ryanstryin posted a graphic photo showing one of the victims lying in the street, it prompted a lot of reflection from both the mainstream media and the public over whether it’s appropriate to publish or share such images. We’ve had these arguments since the advent of photography – in times of war, in times of peace – on whether to publish photos of the dead and wounded or withhold them out of respect for the victims and their families. But this was a special kind of wake-up call. The media no longer makes these decisions, now that witnesses have a publishing platform in their pocket. New media commentator and J-school prof Jeff Jarvis got a little hot under the collar defending his own decision to share the photo on his Twitter stream and offers a compelling argument on the side of keeping the news unfiltered. The point is, if you click this hyperlink showing a victim with blood streaming down the sidewalk (republished here by Slate), you’ve already been forewarned by the linked words. Since mainstream media still have the broadest reach, they will continue to find themselves at the center of this debate, but the audience is going to find it increasingly difficult to avoid such material. The decision will be not one for the “broadcaster” on whether to share, but a personal one on whether to click.

Streaming the world 60 seconds at a time. The Wall Street Journal is now asking its reporters to file microvideo reports using the social media video platform Tout. They’re calling it WorldStream. From Tampa to Syria, you can see snippets of life, the news, and everything else a reporter can capture with a mobile phone camera. A first dive leaves me with the impression that much, much work has yet to be done before WSJ’s WorldStream can be called a mature product. Rebels relaxing in a mosque in Syria might have been portrayed better with a photo, for instance. Thirty seconds watching a pan of the empty delegate center in Tampa would have been better spent reading an actual story about the convention. And I can’t help but wonder what you can expect to get out of a 60-second interview with a pol – the format seems more suited to TMZ celeb shots and gotcha journalism. It will be interesting to see how the service evolves. For now, my main impression is that we’re looking at the news equivalent of Romantic fragment poems – Coleridge’s “Kubla Kahn” or Keats’ “Hyperion.” They may work artistically, but are story fragments really the best approach for an industry devoted to informing and enlightening its audience?

Social media is bull$#!t. Or so says B.J. Mendelson in the title of his new book. The former social media marketer and contributor to Mashable boosts his own contrarian view after serving the industry for years. Among some of the more common precepts of online journalism Mendelson disputes: the all-importance of pageviews, that Facebook really has 800 million users, and that we’ve learned much new about Internet marketing since Dale Carnegie’s “How to Win Friends and Influence People.” He tells journalist Ernie Smith that the biggest BS thing about social media is “the concept that what’s happening on these very different platforms, with their comparatively small and different audiences, has resonance with what’s happening with the rest of us. This false hope we’re giving people, which not coincidentally popped up around the same time the economy cratered. People needed something to believe in, and selfish and greedy marketers were ready to give that to them in the package of the myth of social media.” Incidentally, the interview is a nice display of what you can do with Jux, yet another platform for quick blogging.

The problem with open data. Is there one? Some interesting conversations on the topic this week. One started when the White House announced the selection of its “Innovation Fellows,” members of the private and nonprofit sectors and academia whose job it will be to help develop five government programs, including one on open data. That announcement sparked some backlash from conservative commentators, including Michelle Malkin, who wondered whether this isn’t really just a waste of taxpayer money. Open government reporter Alex Howard captured some of that debate, which unfolded in the social media sphere. Meanwhile, techPresident’s David Eaves reported on how a government spending scandal uncovered in the U.K. with the help of an open data project raises as many questions about how government collects and reports its data as it does about the suspect spending. So, what do you do if the government’s databases are poorly coded or managed – how do we get the government to change? And even if you discover these remarkable stories with the aid of open data sources, does it make it any easier to act? More questions like these are sure to present themselves as data journalism flowers into a discipline in its own right.

Another decade of the Internet. I leave you with a fun look back at how much the Internet has changed in the past 10 years, courtesy of this Mashable infographic. Enjoy.