Challenges Posed by Network Multidimensionality in the Digital Age

Yochai Benkler

Reported by: Nina O’Brien, Allie Noyes & Lauren Frank

Yochai Benkler is the Berkman Professor of Entrepreneurial Legal Studies at Harvard, and faculty co-director of the Berkman Center for Internet and Society. Before joining the faculty at Harvard Law School, he was Joseph M. Field ’55 Professor of Law at Yale. He writes about the Internet and the emergence of networked economy and society, the economic, social, and political roles of commons-based practices in the networked environment, and the emergence of large scale cooperation as a major dimension of social production.

In the second session of the Annenberg Networks Network (ANN) conference, Benkler presented his ideas about networks, power and freedom in the digital age.  He began by reviewing different dimensions of power–including political, industrial organization, cultural, institutional design, technical platform design, and social practices and norms.  He posed a series of questions about how power flows differently in the digital age and then framed his focus as exploring how we use the computer-mediated nature of the networked society to make machine observations more complete and refined as a method of studying social relations.

Benkler addressed the changes in legal and organizational power by providing an example of how the network can be used to emphasize counter power (e.g., free music downloads with request for donations).  He suggested that emerging ecosystems based on things like voluntary sites and fan culture may be able to challenge traditional capitalist power structures and destabilize existing categories of power (e.g., consumers vs. producers).  However, as he builds a possible case for the idea that the internet “democratizes,” he also interjects the major waves of criticism of this idea.

The argument for the idea that the internet democratizes is based on the fact that suddenly anyone and everyone has the power to be a “pamphleteer” and to disseminate information widely.  The first generation critique of this argument is based on the issue of fragmentation.  Although anyone can disseminate information, people are consuming information based on personal beliefs and preferences and are no longer confronted with ideas that challenge their point of view.  The second generation critique of the “internet democratizes” idea is related to the power law distribution of links.  It may be possible for anyone to present information to the world vis a vis the internet; however, the vast majority of the information on the internet is never viewed by a substantial audience.

Finally, Benkler reviewed a number of challenges to studying networks, power and freedom in the digital age.  What is the entity of interest?  Is it the blog or perhaps the individual author?  What is the network of interest?  Do blogs and newspapers get combined into one network?  How is it possible to account for diverse structures within a network (e.g., political networks that differ substantially between the left and right)?  What are the limits of network analysis on these questions?  Is is possible to integrate many different kinds of research into network approaches–like qualitative research, text analysis, offline networked power, money and other power systems, and behavioral/brain sciences?

Discussion

• Capra: Multi-dimensional webs are reminiscent of chaos theory which has succeeded in solving equations with variables and producing compact representations of a system.  Have network theorists thought of defining the space of variables?

Benkler: All of this is at an early stage.  Other people can improve upon it with different skill sets.

• Barzilai-Nahon: With bloggers, do we have replication?  Maybe they are not a mob, but instead merging elites in real time.  What do you think?

Benkler: I have had my share of being overly optimistic about the democratizing effect of the web.  By definition, someone who has time online may be elite.  Does it change from a few hundred or thousand people being able to influence the political system to 2-3 million being able to directly affect and a total of 30 million able to indirectly affect? Maybe now being part of the elite makes you part of 20-30% of population, rather than a fraction of a percent.

Barzilai-Nahon: The real question may not be about the numbers but about what it means.  Maybe the number is not important.

• Latour:  There is a vast amount of information available.  Are we talking about scaling up or compounding so many profiles?  Can we go back and forth?  It seems the advantage should be that we can simultaneously look at the whole and zoom in on specific parts.

Benkler:  We may just be using the term “scaling up” differently.  We are scaling up our ability to make subtle judgments that humans can make without the benefit of a machine.  When we zoom back in, we still get a high-resolution image.

Additional Readings:

Benkler, Y. (2006) The Wealth of Networks: How social production transforms markets and freedom. New Haven: Yale UP.

Moving Technology Inside the Network: Multidimensional Networks in a Pervasive Technology Use

Noshir Contractor, Peter Monge and Paul Leonardi

Reported by: Nina O’Brien, Allie Noyes & Courtney Schultz

Noshir Contractor is the Jane S. & William J. White Professor of Behavioral Sciences in the McCormick School of Engineering & Applied Science, the School of Communication and the Kellogg School of Management at Northwestern University, USA. He is the Director of the Science of Networks in Communities (SONIC) Research Group at Northwestern University. He is investigating factors that lead to the formation, maintenance, and dissolution of dynamically linked social and knowledge networks in a wide variety of contexts including communities of practice in business, translational science and engineering communities, public health networks and virtual worlds.

Peter Monge is Professor of Communication at the Annenberg School for Communication and Professor of Management and Organization at the Marshall School of Business, University of Southern California. His most recent book (with Noshir Contractor) is Theories of Communication Networks. He has published theoretical and research articles on organizational communication networks, evolutionary and ecological theory, collaborative information systems, globalization, and research methods.

Paul Leonardi (Ph.D. Stanford University) is Assistant Professor of Communication Studies, Industrial Engineering and Management Sciences, and (by courtesy) Management and Organizations at Northwestern University where he holds the Breed Junior Chair in Design. Paul’s research explores how information technologies and organizations can be simultaneously designed to enhance one another. His work on these topics cuts across the fields of Organization Studies, Communication Studies, and Information Systems and has been published in leading journals in these fields.

In the first session of the Annenberg Networks Network (ANN) conference, Noshir Contractor began by presenting two opposing points of view.  The first view is that technology is an exogenous variable that can shape networks. The second view is that networks shape technology so that networks, instead, become the exogenous variable. Contractor explains that even though there are two opposing views, they share something in common – they both see technology and networks as distinct entities. It is important that we, as network scholars, attempt to transcend that view so that technology and networks can be brought together. Two theories where briefly discussed which attempt to accomplish this: actor network theory and the sociomaterial approach.

According to Contractor, including technology in networks is a useful step to address many of the challenges researchers currently face. First, in addition to people, nodes could also include things such as documents or computer programs. When the types of nodes change, so do the relationships between these nodes.  People can be friends with each other, but in all likelihood they would not consider themselves friends with their database.  Taken together, this shift would create networks with many different types of nodes and many different types of relationships between those nodes. The question he asks is: Can existing network approaches represent this conceptual shift?

Paul Leonardi then presented a case study that serves as an example of a multidimensional network.  The case focused on engineers who do computer models of crash tests for a car manufacturer.  One engineer developed a computer program to assist with the analysis of the crash test models.  Leonardi described the diverse range of network ties that initially led to the dissemination of the computer program.  Initially when people began to use this new program, they continued to seek advice from other people (i.e., experts) about how to analyze crash test models, but eventually as people became familiar with the program, they started using the program itself as an “expert” in the process of figuring out how to analyze crash test models.  The technology actually became a node in the network.  If it was analyzed only as an influence on the network of human relationships, the nature of the network would be distorted.  The multidimensional perspective allows quantitative network methods to more accurately explore the kinds of complex relationships that are typically left to qualitative ethnographic research.

Discussion:

• Wendy Hall: Consider Implications, for example Google Buzz automatically generated networks for individuals based on their  frequency of email with contacts. This has had some terrible consequences and has been a real privacy problem — they have had to scale it back.
  • The Google Buss issue raises a bigger question: who gets to set the rules? Is the policy an opt-out or an opt-in?
• John Taplin: Question for Paul: before JWIN was introduced there were people identified as experts who were called upon, and they slowly become replaced with technology and the network. What happens to the social relations of those people who used to be the go-to experts
  • Paul Leonardi: Response to JT: expertise isn’t always singular — people may have multiple expertise. As well engineers were often getting dumb questions or practical questions (i.e. how do I use this tool in this context) rather than substantive questions about the “why.”
• Ernie Wilson: How would the model respond if instead of losing connections in one sphere and picking it up in others, we consider the ways in which power changes over time, eg, does individual status change over time?
• Karine Barzilai-Nahon: Not sure this model does account for power relations — it illuminates different relationships, but doesn’t really get at power, per se
• Woody Powell: This case study is lovely because it clearly demonstrates the logic, but scaling up may be premature — we want the specific case to illuminate whether relations change “on the ground”
• David Graywell: Why is technology here represented as a node rather then a linking connetion? What is the value of separating them out as nodes?
  • Response: consider biological networks: it is often useful to distinguish between patterns of organizations and biological structures. Structure os an embodiment of particular relations
  • Yochai Benkler: My take was that this was about working with general purpose tools for dealing with theoretically generated claims about human relations
  • Michael Macy: To the question of whether the nodes have to be motivated or if they can be artifacts, CMC research demonstrates why. A nice metaphor is of two people crashing their cars. That interaction is mediated by the materials of the car they are driving
• Carter Butts: To describe things completely you need lots of variables, but there’s a danger of being overrun by the proffered complexity. a theoretical challenge is to have the development of the models keep pace with theoretical objects. There is a limit on the utility of additional complexity.
• Lada Adamic: How can we do different things with the data than we were able to do before, substantively? Or do we have to reduce things down to crunch the data anyway, in which case we would be comparing more of the same kinds of things?
• Response: Noshir Contractor: in this case the JT node becomes more central, and here it is obvious in a way that would not otherwise be as clear. You can specify the logic of attraction. The real advantage is that you can see those configurations and determine if they are more frequent than random.
• Response to the power questions raised by Wilson and Taplin: marginalization in the network is a distinct possibility — though those marginalized people would show up, perhaps as isolates, so the model does capture that to some extent.
• Paul Leonardi: How can we map these relationships? If you’re interested in power, you can ask those questions too — the advantage of the ethnographic sample is that we have the opportunity to learn what is important to the individuals in the network. That drove the decision about how to model it. That is a really important concern in scaling up to larger networks — how can you keep that kind of specificity alive?
• Is a certain measure an indicator of power,  or is power a relation?

Additional Readings:

Monge, P. R., & Contractor, N. S. (2003). Theories of communication networks. Oxford: Oxford UP.

Leonardi, P.M., & Bailey, D. (2008). Transformational Technologies and the Creation of New Work Practices: Making Implicit Knowledge Explicit in Task-based Offshoring. MIS Quarterly, Vol. 32, No. 2, pp. 159-176.

Ever wonder how news stories are connected? Now there is a way to visualize them -  as social networks. I just discovered this interesting little tool to visualize daily events:

From News Dots:

“Like Kevin Bacon’s co-stars, topics in the news are all connected by degrees of separation. To examine how every story fits together, News Dots visualizes the most recent topics in the news as a giant social network. Subjects—represented by the circles below—are connected to one another if they appear together in at least two stories, and the size of the dot is proportional to the total number of times the subject is mentioned.”

Noshir Contractor, the Jane S. & William J. White Professor of Behavioral Sciences at Northwestern University, talks about his research on social networks. Nosh is the director of the SONIC network research center, which has partnered with ANN to study scientific collaboration in virtual teams.

(via the Center for Internet Research)

The Department of Defense’s TISC (the Transnational Information Sharing Cooperation) has been a central communication tool for relief efforts in Haiti:

From the article:

“The system is designed to be as simple as possible, and is as easy to use as a site like Facebook, says Ty Wooldridge of the U.S. Pacific Command in Hawaii. It uses file-sharing applications, wikis, blogs, and calendaring tools, among other things, to coordinate information and action among people, no matter where they are. Though there are obvious military implications to that kind of network, its first battlefield test is ongoing, on the ground in Haiti.

Without another way of collaborating, the TISC platform has become one of the de facto standards for communication among the relief effort in Haiti.There are more than 1700 different users in Haiti, most of them relief organizations of various size and specialty looking for how to get involved, and to coordinate efforts to maximize results. It’s operating on a larger scale than DISA had originally planned, but it’s scaling well, says Jean Dumay, one of DISA’s leads on the TISC project. “The test came early, and it became very real, but we were ready for it.”

“Distinguishing Influence Based Contagion from Homophily Driven Diffusion in Dynamic Networks” PNAS

Sinan Aral, Lev Muchnik andArun Sundararajan

ABSTRACT:

“Node characteristics and behaviors are often correlated with the structure of social networks over time. While evidence of this type of assortative mixing and temporal clustering of behaviors among linked nodes is used to support claims of peer influence and social contagion in networks, homophily may also explain such evidence. Here we develop a dynamic matched sample estimation framework to distinguish influence and homophily effects in dynamic networks, and we apply this framework to a global instant messaging network of 27.4 million users, using data on the day-by-day adoption of a mobile service application and users’ longitudinal behavioral, demographic, and geographic data. We find that previous methods overestimate peer influence in product adoption decisions in this network by 300–700%, and that homophily explains 50% of the perceived behavioral contagion. These findings and methods are essential to both our understanding of the mechanisms that drive contagions in networks and our knowledge of how to propagate or combat them in domains as diverse as epidemiology, marketing, development economics, and public health.”

Testing the famous “sacred canopy” argument in a social network, this article seems like a fascinating read.

Author: Matthew E., Brashears

Source: Social Networks, In Press, Corrected Proof, Available online 22 January 2010 (URL)

Abstract: This article evaluates the Durkheim/Berger argument that integration in a network of co-religionists protects against anomia. The 1985 General Social Survey network instrument is used to evaluate the effect of integration on anomia and the probability of unhappiness. Results indicate that contact with religiously homogeneous others paired with personal religious belief reduces anomia and the likelihood of unhappiness. Additionally, while ego/alter closeness is important, alter/alter closeness is not. These results suggest that individuals benefit from religious association more so than religious community. Additional analyses indicate that these results are unlikely to be due to homophily.