This week we are featuring our second installation of a series of interviews showcasing individuals in the 2013-2014 Berkman community. Conducted by our 2013 summer Berkterns, the mini-series highlights the unique and multidisciplinary group of people within the Berkman community exploring the many dimensions of cyberspace. Our second set of interviews covers Shane Greenstein's passion to deconstruct the economics behind Wikipedia, Niva Elkin-Koren's research into the political force of online crowds, and Amy Johnson's investigation of online parody in the face of censorship.
Interested in joining the Berkman Center’s community in 2014-2015? We’re currently accepting fellowship applications -
read more here.
interviewed in summer 2013 by Berktern Leonid Grinberg
Shane Greenstein and I met over lunch at Hi-Rise, a small hipster coffee shop a block away from the Berkman Center. Dr. Greenstein spent the first 30 years of his life in the Bay Area ("staying at home has its perks, of course, because it's so familiar... but then again, it's so familiar"), but he speaks with a hint of a Chicago accent, which he acquired, I assume, over the more than two decades of teaching around there: first at the University of Illinois and later at Northwestern. (He currently holds the Chair in Information Technology at the Kellog School of Management.) His background, he told me, is in industrial economics— examining markets and firms, investigating why some firms do better than others, trying to understand whether competitive markets produce better outcomes than non-competitive ones. "Boring stuff, really," he summarized casually. But it led him to thinking about the telecommunications industry just as the Internet was coming to the fore.
What people don't realize, he explained, is that it takes a very long time for anyone to really make money off of a new industry like the Internet. "That's one of history's oldest lessons," he said with some vigor. An easy comparison, he says, is the railroads in the 1850s. In much the same way as railroads transformed whole industries, the Internet completely changed the world. Yet, in its early stages, it's simply too fast and too unpredictable to reliably make money off of it. Dr. Greenstein expects that we are still in the very early stages when few people will actually be able to build sustainable firms with predictable success. "The economy needs time to adjust."
Dr. Greenstein is working on a book on this subject, entitled “Innovation from the Edges: The Economics of Creating the Commercial Internet.” Finishing this book is one of his goals for his fellowship at the Berkman Center. Indeed, writing seems to be one of Dr. Greenstein's passions. His CV lists five books, dozens of papers and chapters in collected publications, as well as over 100 essays for a column in IEEE Micro magazine. Another paper he is working on concerns the fall of Encyclopedia Brittanica. He first brought it up when I asked him how making money off the Internet relates to bias in Wikipedia (a topic the Berkman press release said he’ll be spending his time thinking about). He explained that there's a popular misperception that Encyclopedia Brittanica was defeated by Wikipedia because the people running it were somehow short-sighted, or didn't understand the direction the industry was heading. "And that's just wrong," he says. "Many of the facts are simply wrong." First off, Encyclopedia Brittanica didn't lose to Wikipedia, it lost to Encarta. (I admitted with embarrassment that I hadn't even remembered Encarta until he brought it up.) But more importantly, the executives knew exactly what was going on and, Dr. Greenstein asserts, did everything they should have done. "Everything you could have told them to do with the knowledge they had, they did right." The world was simply moving away from them.
Dr. Greenstein has been working on this paper for several years and says that he had interviewed dozens of people from both Encyclopedia Brittanica and Encarta to try to understand their side of the story. As a scholar, he had been interested in encyclopedias for a long time, and it's that interest that brought him to think about Wikipedia.
Wikipedia, he explained, is a fascinating subject for an economist, in part because it's a simple, objective demonstration that existing models of the economy just aren't good enough. "It's obviously worth something, it's displaced a billion-dollar industry," he points out. "Yet GDP says it's worth zero. So obviously we don't have everything right." It's a simple observation, and one that connects back to an issue near and dear to the Berkman Center's heart. Why do people edit Wikipedia? Who edits it? And is the result any good? The economic perspective, however, gives it a new and interesting twist. How do we actually quantify its usefulness to society?
This is where the study of bias comes in. Dr. Greenstein recites "Linus's Law," which is named after Linus Torvalds, the original writer of the Linux kernel: "Given enough eyeballs, all bugs are shallow." It's a law that, empirically, has held very true in open source software projects, but it's much harder to evaluate in something like an encyclopedia (in part because it's just so much harder to tell what a "bug" is). Nonetheless, we can evaluate the purported claim of neutrality using relatively standard methods—for example, examining which sources are cited and tracking the prevalence of “reversion wars,” a phenomenon in which a user edits a page, another user reverts the change, the first user reverts the reverts, and so on, back and forth. Some of the results match expectations: for example, reversion wars tend to be common only with highly controversial articles, and the perceived degree of bias tends to be much lower in objective articles about scientific or technical topics, as compared to political ones. But some results are more surprising: for example, even in the technical topics, most of the cited sources tend to be popular news articles, such as The New York Times. Citing scholarly or medical journals is quite rare, even if the topic is academic.
Dr. Greenstein teaches these results to some of his MBA students. Wikipedia, he has found, is a perfect way to explain concepts like Linus's Law and other topics relating to the open source world when presenting to a non-technical audience. Talking about code is quite difficult because it can be hard for a non-programmer to understand that different people might have widely different interpretations of the same piece of code. And open source in general is really hard to teach to MBAs and executive students. "They just want to ask if they should let their employees do it," Dr. Greenstein laments. "And that's not the right question to be asking."
Dr. Greenstein will be splitting his time between his home in Illinois and working in Cambridge, where he will be conducting research, working on his book, and teaching at the Harvard Business School. The father of four children, he said it will be hard not seeing them, even though he will be home on most weekends.
But what might have been a tough pill to swallow not ten years ago is made easier with modern technology like Skype. Dr. Greenstein says that he says he plans to use Skype regularly, particularly to see his youngest son. Paraphrasing Woody Allen, he explains, "90% of being a good parent is just showing up. Most of the time you show up and they don't need you, but you really want to be there the one time they do."
interviewed in summer 2013 by Berktern Alysa Batzios
What are you working on now? What would you like to explore in the coming year?
NE: My research these days focuses on the new types of threats to freedom in the digital ecosystem and the legal mechanisms to address them. We live in time of transition. The past two decades have seen the rise of distributed systems for generating and distributing information and promised a more participatory environment, promoting innovation, democracy and justice. This environment is now rapidly changing, creating new types of challenges for freedom online.
I am currently studying crowd management. I am particularly interested in the rise of online crowds as a political force, which destabilizes governments and commercial interests and triggers strategic intervention. Strategic interventions and manipulations by governments and commercial players (often in collaboration, using similar measures) may include ongoing surveillance, astroturfing, and the use of social media analytics. Crowd management strategies may are challenging our democratic baseline. To address these challenges we need to configure new checks and balances for these types of threats, articulate digital rights and re-negotiate our constitutional rights.
A somewhat related study focuses on the changing nature of online intermediaries, and the new convergence of control over content, access and end-users. The multiple roles of online mega platforms may involve intermediaries in a conflict of interests when they engage in online enforcement. In this context I also run a study on actual practices of notice and takedown.
Finally, I study user rights under copyright law. The rights of online users are increasingly under siege due to the shift to digital and the expansion of copyright protection. Against this background, I seek to offer a legal framework that moves beyond exceptions to copyright view permissible uses as rights. A comprehensive approach to user rights may provide the theoretical and doctrinal bases for the dynamic development of these rights and for strengthening their legal status.
What do you see as the most pertinent legal issues regarding access and use of information?
NE: Questions of access and use of information should be understood in view of civil liberties. In this context, there are at least three major shifts that I believe we should worry about:
One is the rise of an infrastructure of control: we see more filtering, blocking and monitoring for the purpose of copyright enforcement (e.g., Content ID, DPI), but also for law enforcement in general as well as advertising and sales. These practices are often non-transparent and are not subject to legal scrutiny.
Another issue is the weakening power of users at the network ends. This is linked to the rise of online mega platforms, cloud computing, the shift in content distribution from products to services and the dominance of private ordering. Overall, these developments keep the ends with less control over the type of information that could be shared and accessed, and especially over the terms for doing so.
Finally, we should be worry about the collaboration between governments and online intermediaries in law enforcement tasks, without sufficient legal oversight (what my co-author Michael Birnhack and I named a decade ago the Invisible Handshake). This government-market cooperation is becoming an important tool of governance and it is most likely here to stay.
Overall, those shifts put us at a turning point: this is not simply about access to information, it's actually about freedom, and we currently lack the legal safeguards for securing it.
In your opinion, does copyright law impede or facilitate the acquisition of knowledge?
NE: As the gap between copyright law and the actual needs in the digital ecosystem widen, copyright is effectively slowing down innovation and stifle access to information. A particular concern is the way copyright is invoked to protect existing institutional structures and maintain control over the production and distribution of cultural resources. For example, the legal battles of the broadcast television industry against streaming companies such as Areo, or the use of copyright by book publishers to maintain their hold over digital publishing in their legal battle against Google Books which was concluded in a confidential settlement agreement.
Have you detected changes in the application of fair use provisions with respect to educational institutions?
NE: Educational use is a typical example which demonstrate the role of users in promoting copyright goals: promoting progress requires learning, training, the opportunity to absorb informational goods, and to interact with it.
Educational use require that access to learning materials remains affordable and that licensing fees do not prevent learning; it should enable professors to use materials spontaneously for educational purposes and respond to matters in a timely manner without having to clear the rights months prior to the beginning of classes; learning involves more than simple access, it also entails interaction with copyrighted materials, and therefore the freedom to use (explore, tinker, transform) and edit materials before making them available to students.
Unfortunately fair use analysis is still far from securing educational use. After years of very little litigation on education use – and a very narrow interpretation of what is permissible – there were recently a couple of important decisions on educational use. One case is the lawsuit of major academic publishers against Georgia State University regarding the use of copyright materials in electronic reserve system Cambridge University Press v. Becker (2012)).
There are some good news and bad news here: The court held that the use of small excepts of scholarly materials in e-reserve might be fair use. The bad news is that the court held that the availability of a license may tilt against fair use: when permission is available --- at a reasonable price, and in a convenient format – unpaid use weighs heavily against fair use. For a use to be considered fair, it has to be the lowest of either one chapter or 10% of a book. The message to publishers is clear: offer licenses for each and every use and also, edit books in smaller chapters.
Yet, the main problem with fair use for educational purposes is the high level of uncertainty. The absence of any quantitative guidelines might be polarizing for risk averse educational institutions. In the GSU case the court held that even the stated amounts (1 chapter or 10% ) are subject to final determination using the multifactor analysis. This forces educational institutions to run the multifactor analysis for each of the items which are made available to students. It took the court 350 pages of legal analysis to reach a decision on the 75 of the 99 claims of infringements submitted by the plaintiffs. This is not a feasible exercise for non-layers who are asked to handle thousands of learning items on an ongoing basis.
What do you hope to achieve through your relationship with the Berkman Center?
NE: I look forward to engaging in stimulating conversations and to the opportunity to brainstorm and try out some new ideas with such an impressive group of interesting scholars and activists.
interviewed in summer 2013 by Berktern Keisuke Otsubo
What is your main research which you working in Japan? And why does it have to be in Japan?
AJ: In some ways this is a tricky question to answer. I’m interning at Twitter Japan this summer. My project there is researching how Twitter has been used in crises and natural disasters across the globe. The 3.11 triple disaster was, at the time, the most digitally recorded disaster ever. (Hurricane Sandy in 2012 may have generated more digital records; I haven’t seen any estimates.)
Taking a step back, I’m interested in giant questions about how communities form with media technologies involved. In the past we tended to define communities in terms of language and location. Does this make sense now that media interpenetrate the globe? How do elements like identity and membership, community and intimacy, manifest when we include the complexity of media use? Performance has been a dominant trope to explain and explore many of these elements. Should we consider things, instead, through the lens of animation? To investigate these questions, I’m examining how different cultures use Twitter, how Twitter as a company that has its own culture and cultural background adapts to different cultures, and how cultures that are arguably native to Twitter (or the online world more broadly) mix with these other two. That was probably a horribly opaque way of describing it. Where do the lines of a community lie? Or is the very concept of a bounded community outdated and in need or re-envisioning?
Right now there’s a heavy push to study all sorts of events through big data. Big data is certainly intriguing; our ability to study phenomena on such a large scale, with so many discrete data points, has definitely expanded. But at the same time, the little data is being overlooked. (I think I may have just invented a new term!) The little data—which isn’t really all that little by most standards—relies on qualitative rather than quantitative approaches. I’m interested in this little data, in qualitative approaches to a field that’s wallowing in big data. Everything from interviews to ethnography to discourse analysis to meta/pragmatics and beyond. (I should say, too, that though I focus on qualitative methods, I also use quantitative methods like social network analysis and strongly believe a mixed approach is the best.)
Digital records, of course, are often accessible across the globe. But understanding the richer context of those records demands the little data. So I’m at Twitter Japan this summer not just to speak to the people involved in dealing with the Tohoku disaster (in terms of the Twitter side), but to understand the context in which they work.
Why did you choose Japan? Why is Japanese language and Japan was special for you?
AJ: Japanese is fascinating. You’re probably so used to it that the mixture of writing systems doesn’t even seem interesting to you, but it is truly a linguistic marvel. As you know, in Japanese there are four different writing systems in use (hiragana, katakana, kanji, and romaji). Some items are always written with one system; for example, inflections on verbs are always written in hiragana (e.g., ??? ?). In other cases, the same word may be written in hiragana or katakana or kanji. The choice depends on a number of primarily sociolinguistic factors. This is a huge, huge topic, but suffice it to say that sociality is thus embedded in Japanese writing systems in a visible way different from any other language I can think of. It’s beautiful. And fascinating.
One of the big problems that linguistics, sociolinguistics, and linguistic anthropology share is an over privileging of speech and a consequent neglect of writing. The study of writing has long been informally allocated to the field of literature. However, computer-mediated communication (CMC) provides the opportunity to examine the sociality of writing with fresh eyes. Thinking about English, Japanese, and Arabic together—and environments in which they mix—offers an entirely new way to think about writing, language, and communication. It doesn’t have to be these three languages, but these three are particularly good candidates. English is a global lingua franca at a level never previously seen; it continues to dominate in tech spheres. Although Japanese and Arabic are very different from each other and English, both languages are marked by intriguing complexities. In particular, the writing systems and registers of Japanese and the multiglossic spectrum of Arabic. I’m fascinated by the fact that telegrams were mainly written in katakana. Early CMC in Japanese was also in katakana—and then at some point it shifted to what we know today. Why? How did that shift happen and who was involved? I don’t know, but I’d love to!
Separate from the elements that thrill the linguistic nerd in me, the trio of languages have very different histories with Twitter. English has always been the most common language used on Twitter (not surprising given that it’s from a US company). However, use in Japan/Japanese has been high basically since Twitter was introduced, with Japanese the second most common language on Twitter. But because of its intertwined writing systems, Japanese has incredible density variance. That is, you could write a tweet that packs far, far more words into 140 characters than in English. Or you could write one that’s pretty similar. But you have the option. So tweets in Japanese have a very different spectrum of possibilities. Arabic, of course, broke Twitter’s top ten languages in 2011 after a year of incredible growth in conjunction with the Arab Spring. It has its own interesting history with CMC, in part because neither ligatures nor right-to-left orientation was supported for a long time. As a result, there’s a special form of Arabic that originated in CMC, called Arabizi, that uses roman characters and breaks all sorts of rules. In terms of languages, it’s a disruptive technology. :)
What are your Ph.D. major and profession? Does it relate to your current stay and research in Japan?
AJ: I’m a linguistic and media anthropologist in MIT’s HASTS PhD program. (HASTS = History, Anthropology, and Science, Technology & Society). This fall I’ll be entering my third year in the program. My research this summer is pre-dissertation research—it’s laying the foundation for further work I want to do over the next two years or so.
What is your main research theme? Please tell me more about “what happens when online parody is taken (too) seriously"?
AJ: Parody (and humor) serves as a lens through which the giant questions that interest me can be focused. Parody is a truly fascinating form of communication, because it relies on context. It’s the antithesis of referentiality. In order to understand and “get” parody, people have to recognize not just what is in front of them, but its relationship to something that may be nowhere in sight. Context is critical. This makes parody an excellent test case in a number of different ways. First, who gets particular parodies? (“Get” here both in terms of who can access them as well as in terms of who understands them.) And who doesn’t? Parodies thus allow us to trace the shapes of publics and communities. Because parodies often face censorship—or attempted censorship—legal cases focused around parodies also tend to make manifest media ideologies: lawyers and judges and expert witnesses are forced to interpret parodies, leading to revealing testimony.
The question from my bio that you quote above seeks to pin down three different “serious” uses/responses to online parody: 1. The use of parody for social critique, whether that’s direct political or oblique critique (e.g., Twitter parody accounts of politicians and mock Amazon reviews). In some cases such parody is easily recognizable as activism. Online, these forms of parody are often collaborations that break the traditional two tiers of producers and receivers. How does this affect their creation and consequences? 2. The censorship of parody. Why are some online parodies censored in some circumstances? Parody demands an individual response, examination on a case by case basis, which is part of what makes it so fascinating. Intertwined with discussions of parody are questions of authorship and authority and sanctity and privacy and security—even how we experience reality. How do we distinguish between the parody and the original? How does media affect/facilitate/modify the intensity of our experience? Court records of censorship trials in Japan reveal complex arguments about whether literary, kinetic, and multiple modes of media more vividly represent reality—and thus are more dangerous. In Egypt, Bassem Youssef—whose satirical work was first broadcast via YouTube during the Arab Spring—has battled charges that he is disrupting the peace. 3. Finally, misunderstood or mistaken parody. There are numerous examples of parody news reports that have been mistaken by one country or another, one news station or another as actual news. The parody has gone unrecognized. In the US, we probably hear about this most often with regard to the Onion and foreign countries like China and Iran. Why do some parodies go unrecognized? And what does the discourse about this mistaken parody—usually by other gleeful news agencies—reveal about connections between communication, understanding, and power?
Can you specify more about your and Berkman Center relationship? Because you are in MIT that already have strong relationship with Berkman (e.g. Cyberscholar Project), but why do you choose to be a fellow at Berkman Center?
AJ: This past year I’ve been only peripherally involved with the Berkman Center. But even just these experiences have whetted my appetite for more. The Cyberscholars Project was something that I became involved in through several friends in the Center for Civic Media, which is collaboration between the Media Lab and the Comparative Media Studies Program (now Comparative Media Studies & Writing). I have been incredibly lucky to have two bases at MIT filled with interesting, creative people: the HASTS program and then the Center for Civic Media, my home away from home. Every experience I’ve had with the Berkman Center so far suggests that it, too, is a focal point for people I want to talk with, argue with, collaborate with. Both the questions asked and the projects undertaken interest me, at a scale ranging from the work on narcotweets by Andrés Monroy-Hernández and Panagiotis Metaxas to Creative Commons and the Open Net Initiative’s work on online censorship. I’m particularly looking forward to the weekly Tuesday meetings and the structure they provide—I’m excited for more sustained engagement with other fellows and Berkman folks.
According to your world creation in multiple linguistic environments research, is there any features or special difference between English language and Japanese language?
AJ: Still working on that one! Give me a year or two and I’ll get back to you.
This post is part of a current series of interviews with members of the Berkman Community. Previous entries: Young and Livingstone.