GigaNet Symposium

5 December 2016 - A Pre-event on Other in Guadalajara, Mexico

Also available in: English
Full Session Transcript

    >> MARIANNE FRANKLIN:  Good morning.  Can you hear me?
    Hello!  Can you hear me?  Have we got mics?  Yeah?
    Okay, Gracias.  Good morning, everybody.  Welcome to the GigaNet Annual Symposium 2016.  It is in fact, I believe, our 11th Symposium.  And we are officially ten years old.  My name is Marianne Franklin.  I'm Chair of the GigaNet.  And I would like to welcome you here.
    We have some issues with people still waiting in line.  So we've decided in the interests of time to begin.  So our Programme Chair, Daniel Oppermann, will say a few brief words about the programme and will be the moderator until our moderator, Alejandro Pisanty arrives.
    We are excited.  We have a lot of new papers and faces and some very established and senior scholars with us already in the room.  We are looking forward to learning a lot and having some good, fruitful discussion today.  Daniel will update you now on how the day will proceed.  Thank you very much and welcome!
    >> DANIEL OPPERMANN:  Thank you very much, Marianne.  Good morning, everybody.  Also from my side it is very nice to see you here at the GigaNet Symposium.
    As Marianne said we are still waiting for some more people to get in.  But I can already say that the table is open for everybody.  So if you want to sit on the side, you can sit on the side.  You can also join the table.  So we are closer to each other and maybe could improve interaction later.
    So as you see, we have a printed programme on the tables and on some of the Chairs.  So you can find out that we have four panels today.  We are going to start in a few minutes with the first panel on Participation, Transparency & Responsibility.  Our speakers are already here.  We are still waiting for the moderator, but we are starting already.  So he is going to come in soon.
    After that we are going to have a short break of a few minutes.  Then we go over to the second panel, which is on Legal Challenges, which is going to be a round table.  We will have a round table discussion.
    After that we have lunch break, which this year is quite extensive.  We have like 90 minutes -- actually, before it was like two hours but we cut it down to 90 minutes so we have more time for the afternoon session.  We start at 2:30 with our third panel on actors and policies in Internet Governance.  We have another short break.  And our final fourth session is on emerging economies and crisis.
    After that, we will have the final session, the business meeting, which is focusing on GigaNet as an organisation, on our plans and what we want to do in the future.  We want to have your feedback as well on your experiences, on what you want to bring in maybe, how you could participate or any suggestions are welcome in the last session.
    After the last session, of course, we have a little party as well.  The GigaNet reception.  And on the programme you can find the address already.  It is going to be in a restaurant called Milo, which is quite near the Hotel Grande Fiesta Americano Guadalajara Country Club, the hotel where we had our session yesterday.  Some of you were there already.  You know the neighborhood is quite near that hotel.
    So let me welcome everybody here and I would suggest we start now with our first panel.  One comment?
    >> MARIANNE FRANKLIN:  I'm sorry, I forgot to say.  Apologies in advance.  I have to leave briefly.  I will be back.  If Daniel later has to leave briefly, he will be back.  We have GigaNet business in other rooms.  We are not deserting, it is not that we don't want to hear what you are saying.
    Please, join the table on the other side.  It would be great.  Please come up and join the table.  The table is open to everyone.  Thank you.
    >> DANIEL OPPERMANN:  Thanks, Marianne.  Let's start with the first panel.  Rachel Pollack is here.  Could you please come here?  For the discussion later it is important that if you want to participate in the discussion, you can use the microphone.  So every microphone, there is this large button on the top if you want to use the microphone.  You press it, you speak and press it again to mute it.  Thank you.
    (Pause.)
    >> RACHEL POLLACK:  Hi.  Governing, everyone.  My name is Rachel pollock.  I will be presenting today a paper entitled opening the black box in search of algorithmic transparency.
    I would like first to preface by saying some of you may know, I work at you know he is company in the day job but here as a member of the ICANN consistency.  The paper is based on research at the Oxford Internet Institute last year.
    So my talk will follow, I'll cover a few areas.  First starting with some background and a brief review of the literature.
    And then discussing the research questions and methods that I used.  I'll follow that with a discussion of my findings, including a quick actor mapping of positions.  Discussion of the ideas and interests driving those positions.  My conclusions related to the impact and the types of strategies that were used.
    Finally some recommendations and an indication of areas for future work.
    So why study the regulation of search engine algorithms?  This is the area, the intersection of two major areas:  Research and public policy debate.  On the one hand we have search engines which have been written about extensively over the last 15 years, as they have been called the epicenter of the early web, gateways to information and knowledge and new gatekeepers.  There have been books about the search engines, society and search has a database of intentions, giving insights into human behavior.
    At the same time more recently there has been increased interest in the role of algorithms in society.  And algorithms briefly we can describe as a set of steps used in a process to solve some problem, usually by computers.  That's a simple definition.  There has been comparisons to a cooking recipe, and the steps needed.  But that's also seen as a bit over simplified.
    Around this term of algorithms there has been a lot of mystique developed, some refer to it as the algorithmic turn in research.  And especially concerning the opacity and proprietary algorithms.  So this has been pioneered by the work of Frank Pasquale, the area of black box, search, reputation management and finance.
    So looking at these two areas, this is actually where search is the area where research has come the furthest in terms algorithmic transparency.  Going back to the year 2000 and the work of Helen Nissenbaum and the recognition that search is not neutral and has inherent biases in terms of the sites that it promotes and reinforcing existing hierarchies.  They called for transparency of search engine algorithms or the ranking algorithms.
    Since then there have been other calls from Pasquale and others about how to increase algorithmic transparency and search.  On the other hand, there are challenges to that, including if algorithms are made transparent, first it hurts the business model of search engine companies.  That is their competitive advantage and what is called, often referred to as their secret sauce.
    At the same time it also could hurt the quality of search results because it opens up to spammers, the risk -- sorry, of spammers gaming the system.
    So while there has been a lot of academic and theoretical discussion, to my knowledge there have not been studied about how this has played out in practice, in policy settings.  That's what my research attempted to do.  To give you just a few examples from the last few years, of policy developments that could have an impact, in France there is a new law on digital Republic.  That includes sections related to a concept called platform loyalty or platform fairness, which is referring to both fairness towards consumers and also to other businesses.  And also it contains a provision of algorithmic transparency in the public sector algorithms.
    There was a Senate amendment in France last spring, spring 2015, that was more closely related exactly to this question of algorithmic transparency in search that would have called for greater transparency.  It was ultimately defeated.  I'll go into some of the reasons for that.
    At the European level with the new data protection regulation there are provisions that some have interpreted as the new right to explanation, to receive an explanation of how automated decisions were made and the ability to challenge them.  This also could have an impact.
    There are ongoing competition cases led by the European Commission about Google's practices.  There have been some public statements made, most famously a recent statement by Angela Merkel about the role of search engine and social medium, algorithms and the need for them to be more transparent.
    My key research questions were in three parts.  First I asked how have key actors in Europe and North America positioned themselves on the question of transparency of search algorithms.  I focused on Europe and North America because that's where these debates have gone the furthest.  They also have strong traditions or regulations and economic policy and competition policy and data protection policy.  So I focused on these two regions.
    Secondly, what underlying beliefs and interests can explain these positions?  Finally, what strategies have the various actors used to advocate for their positions and what impact have they had on the efforts for or against greater transparency on the search engine algorithms?
    The methods I used were qualitative.  I conducted expert interviews with about 20 actors from across civil society, academia, tech companies, regulators and intergovernmental bodies across Europe and North America.
    I also analyzed textual material from proposed legislation, governmental debates and media coverage and conducted a thematic analysis.
    My findings, I can sort of summarize in the three, following the three research questions.  First, to understand what positions actors have taken, the major groups that I covered were civil society, Google, its industry competitors, and primarily in vertical search.  So not -- competitors and horizontal search would be conducted by Bing and Dot Dot Go and vertical markets, shopping, price comparison sites.  I looked at governments and intergovernmental organisations.
    And so I didn't try to capture the range of views within these actor groups, rather those who made statements or clearly taken positions or who have had an interest on this question.  And so what I found through this was that many different groups are talking about the need for transparency.  But the reasons for it and what they mean by transparency are often different.  I actually was struck when reading Isadora's paper with the discourse analysis of Cyberpunks and you spoke about freedom was an empty signifier.  In this case transparency is a bit of an empty signifier in the way that each of the groups refers to it.  For example, in parts of civil society, namely the Group Ethic, the Electronic Privacy Information Center in Washington, they call for full transparency essentially publishing the source code of the algorithm.
    Google talks a lot about transparency and points out its efforts, but those are more are in terms of transparency reports of requests that it receives for content take-down or user data.  It also has a blog called Inside Search that talks about updates in very general terms.  They say that they are transparent, but they don't mean in the way of actually publishing the algorithm.
    For some of the industry competitors, namely there is a lobbying body called fair search that is made up of sites like Yelp and Expedia, Trip Advisor and Expedia, they also talk about transparency.  Similarly they are not referring to transparency of actually publishing the algorithm but rather transparency of Google's policies.
    Governments, again it is not a full view, but from the proposed legislation the Senate amendment actually would not have again required publishing the algorithm.  But that was the way it was portrayed in the media and the way that Google also wanted to present it, which I think led to its defeat.
    So to look now at the ideas and interests -- two minutes?  Oh, wow.  Okay, very quickly.  Greater transparency, the main areas were effects on fundamental rates.  That was shared by civil society groups, European institutions, largely there was also especially in France a fear of American technology giants and the kind of the role of Google, the fear of becoming a digital colony of U.S. companies.
    Against greater transparency, there was a recognition of the importance of Intellectual Property, not only by Google but also its competitors.  That actually, I thought, was why the competitors were also against algorithmic transparency because it would apply to their own companies as well.  Then there was wide recognition or agreement across the board that existing laws and regulations were sufficient and that there was no need for a new law specific to algorithms.  Both from the perspective of companies, but also those interested in human rights who said that a new law could actually weaken the position.
    Very quickly, lobbying strategies.  I saw it was primarily lobbying with information and ideas or informational lobbying of companies and organisations commissioning reports from experts to reinforce their existing positions.  I observed a very fragmented landscape of expertise, each referring to a different expert at a different University.
    Just in conclusion, my overall finding was that due to this conflict of interest between Google and its competitors, it seems very unlikely there will be a legal requirement for search engine algorithmic transparency.  It may happen in other areas, but I don't think that it will in search.  Based on this research, I put forward a few recommendations.  Namely:  For some kind of qualified transparency, perhaps given to regulators or self regulatory body but not publishing the full algorithm.  To have more consumer choice and giving Options about whether they want their activity tracked on search.  More education awareness raising.  That's where I will end.  I would very much welcome your feedback.  Thank you for listening.
    (Applause.)
    >> DANIEL OPPERMANN:  Thank you very much, Rachel.  I suggest we continue with the second and third presentation.  After that we will take questions to all three speakers.  The second presentation will be Isadora.  Yeah?
    >> ISADORA HELLEGREN:  Hello, everyone.  It is a great pleasure to be here today to talk about various approaches and aspects on issues of Internet Governance.  My name is Isadora Hellegren.  I recently graduated from McGill University, and I will be presenting the research I did on what I call crypto-discourse.  It is ... sorry, did you hear what I was just saying?
    Should I repeat that?  Okay.  Right.
    I will be presenting my research that I did on crypto-discourse and its implications for our understanding of Internet freedom.  And to do that I will begin with the brief story that happened rather recently that made many headlines.  In February, the FBI and the multinational tech company Apple entered into this public dispute regarding access to data residing in a specific iPhone.  A court order ordered Apple to create software to enable access to this iPhone to protect the content of the phone that belonged to the deceased San Bernardino.  But Appel CEO Tim Cook published a letter publicly opposing the government's request.  Instead of complying with the court order, Cook advocated for the need for strong encryption to protect personal information.
    Apple's response to the court order quickly became a matter of discussion in many media outlets.
    For example, whistleblower Edward Snowden presented Apple as a privacy rights defender:  Apple is defending rights rather than the other way around.  It was written in The Guardian the legal conflict between Apple and the FBI is not even a question about security versus privacy.  Instead, he meant that it is a conflict about legitimacy and he pointed out a democratic deficit in the lack of trust and accountability of the national security agency of the United States.
    In a similar vein, what Colombia Cyber Fellow Silverburg called for the wrongful request as one that requested encryption.  Apple's choice to publish a letter enframing this issue is populism, plain and simple.
    These statements are but of a few examples of the voices that compete in finding the meaning of this dispute concerning law enforcement's ability to access personal data.  The dispute reminded many technology reporters of a series of legal battles that took place in the early 1990s, namely the crypto-are whats where online rights Advocates such as the electronic frontier foundation and the United States Department of Justice disputed the legal status of crypto in court.
    Crypto, I mean is an empty signifier that serves as a battlefield in a larger discourse I have struggle to determine the meaning of Internet freedom.
    So crypto is short for cryptography.  It refers to in its particular meaning to encryption software such as pretty good privacy or new privacy guard GPG.  In the design of this software aims to render the content of online communication illegible to any third parties trying to access it.  But crypto-or encryption software does not have inherent meaning in and of itself, which we can see in a change of legal status over time.  It used to be classified as war material by the U.S. government and by international treaties, which it is not today.  Just last year the United Nations Special Rapporteur on Freedom of Expression presented a multi-stakeholder report in which encryption was described as an essential tool for the protection of human rights online.
    So drawing on the theory of discourse, I conceptualized crypto as empty signifier, a result of inherently political struggles to define meaning.  Empty signifiers are in our sight where we can see these struggles taking place.  Crypto took on possible meanings beyond the features of the specific technology.  In this struggle, it does not take place only in courtrooms but academic journals, conferences, online spaces and in technology magazines.  In efforts to shape the form and function of Internet-specific technologies are not separate from attempts to construct meaning through them.  Media representations of these technologies, practices and their users shape understandings of the Internet.  Depiction of encryption software as a terrorist tool for communication, online file sharing as criminal activity associated with piracy and hackers as villains stealing credit card information communicate what constitutes and what does not constitutes desirable online behavior.
    So Internet Governance debates are both about the technical architecture as well as what (Denardis) called expressions of mediation over societal values such as freedom, Intellectual Property rights related to this architecture.
    Negotiation of this meaning shapes the related policymaking.  So we must therefore also try to understand how various stakeholders construct specific understandings of Internet-specific technologies.
    Drawing on the theory of discourse, I decipher what I refer to as crypto-discourse, a partially fixed construction of meaning that established a relationship between crypto and a negative conception of freedom in relation to the State.
    At the core of this discourse we find Cyberpunks, the Cyberpunks formed as a community when privacy concerns and technology-interested individuals gathered on the electronic Cyberpunk mailing list in 1992 to discuss and develop encryption software in order to protect their own online privacy.  Tech journalists from Wired Magazine perpetuated the work of the Cyberpunk community.  Others reinvigorated their discourse at later stages.
    I mapped events at the protection of crypto related to freedom among cryptographers, hacker activists and technology journalists during a period of 40 years.
    The timeline begins with academic journal Articles from the cryptographic community at MIT and Stanford in context with the U.S. government had the upper hand in defining the meaning of crypto and following the league of the Pentagon Papers showing the role of the government in the Vietnam War, the U.S. experience arising antagonism towards the State.  This is also a time of a rising counter culture where technology interested Members of the whole earth network primarily based in California brought together communities of journalists, hackers and entrepreneurs around the relation between the concepts of technology and freedom.
    And this is the origins of crypto-discourse that set the stage for the crystalisation.
    During the early 1990s, the U.S. government played an essential role in politicizing crypto-Advocates parallel to government efforts to establish the role between computer crime and encryption.  Users from the crypto-community mobilized behind cryptography as a means to achieve political and societal change.
    Drawing on the social antagonism that was already there, the Cyberpunks came together and wrote manifestos that circulated the Internet and come to signify the practices of the entire movement.  These manifestos defined crypto as a force of its own beyond human control, future development that is inevitable.  The use of historical materialism excludes responsibility from social actors to determine its direction and use.
    Any attempts to hinder or regulate the development of crypto-anarchy in this case that they sought is therefore in this representation meaningless.
    The timeline ends with the revitalization period.  This period is characterized by events such as 9/11, globalised anti-terrorism state discourse and harmonisation of policy along with the new uses of encryption.  By new uses the mean the increased, Onion Routing such as the (Tour), mass leaks by Wikileaks and finally Edward Snowden.
    This is used to crystallize discourse in new contexts, extending to journalistic practices and whistle blowing on a global stage.  This new context also globalises social antagonism towards the state that originated in a very specific Anglo-American context.
    Now, how does it function in the crypto-discourse?  I mean that it removes responsibility from democratic states to secure online rights and freedoms as it excludes other possible and positive means of Internet freedom.
    Crypto-discourse presupposes individuals' responsibility to protect themselves and their online communication through technological means from undue interference from state authorities or corporations.
    This understanding does not emphasize the state's responsibility not to use its power in the form of mass surveillance:  Nor does it call for mechanisms that would hold the state accountable if it did abuse its power as a democratically elected government.  This process is not only taking place over time but also over space as it has traveled from a specific context to now encompass many of the world's governments through harmonisation of policy.
    This harmonisation does not take into account differences in legal and political systems such as varying perceptions of the meaning of privacy and the role of government.  Like Jeanette Hofmann said yesterday, what is censorship in the United States is not censorship in Germany.
    And while strengthening individual rights to privacy, crypto-advocates discourse strategies may undermine efforts to construct positive meanings of Internet freedom.
    If the discourse I have struggle taking place in crypto- is indeed over shadowing a legitimate crisis in the United States, then crypto-discourse as currently articulated by advocates should deepen it by further removing responsibility from government.  What is more, crypto-discourse forwards a negative conception of freedom internationally by making individual privacy an individual rather than state responsibility.
    Future encryption policy should therefore seek to take into account national variations in perception of freedom, and consider what should constitute desirable governmental responsibilities in a democracy.
    Thank you.
    (Applause.)
    >> DANIEL OPPERMANN:  Thanks a lot, Isadora.  I'm happy that we have also here now Alejandro Pisanty, the moderator of this session and who made it through the registration.  I guess there were a lot of people outside in the queue.  We are going to continue with the third presentation and after that we start with the discussion.
    >> ALEJANDRO PISANTY:  Thank you.  My name is Alejandro Pisanty.  Welcome to Mexico!  Can you hear me?
    I have two words, one of them is very welcome.  The other is remember, you are 1500 meters above sea level.  For the Europeans, remember that 1500 meters above sea level, you only have ski stations, no cities.  So maybe you are not completely used to moving around at that altitude.  It is the altitude of Denver, for the U.S.  That said, welcome and thanks.  I believe in self moderation so strongly that I actually trusted the session without ...
    (Lost audio.)

    >> BRANDIE NONNECKE:  Okay, is the mic on?  Great.  I can speak directly into the microphone.
    Well, good morning.  I would like to thank GigaNet for organizing today's events.  I have organised many events and I know you put in many hours.  Thank you very much.  I think we will all benefit from today's meeting.  Even looking around the room I see many of my academic idols.  Great, I hope I get to pull them to the side.  I'm Brandie Nonnecke, post-doc for the Center of IT Research in the Interests of Society, headquartered at UC Berkeley, but we appropriate across three of the 10UC campuses.  My collaborator, Dmitry Epstein, was not able to make it to today's event, but I think he is watching right now.  We'll see if he has any comments about my presentation.  Today I will be talking about research we are doing on the ICANN strategy panel on multi-stakeholder innovation, talking about crowdsourcing Internet Governance.  The work that Dmitry and I do, we look at how does the IGF function and look at multi-stakeholderism and how do we define governance and look at the interests connections between these two.
    From the inception and the legitimacy of the multi-stakeholder model has been questioned because some stakeholders may be able to carry more weight.  My dissertation research which feeds into this research looked at the impacts of the East Africa Governance Forum on energy making in the region.  I'm sure all of you know that the IGF is a nonbinding Forum.  However, I have been able to identify with some of my research that the Forums in the East Africa Community, they actually influence actual policy.  So then that begs the question:  Who is participating in those face-to-face forums?  Who has influence in those forums?
    So we are expanding this idea to look at the influence of stakeholders within the face-to-face interactions of the Internet Governance Forum but also how can we use online tools to help create a more inclusive and equitable environment for stakeholders?
    Also I think it is critical that when we discuss the Internet Governance Forum and other multi-stakeholder forums we consider what we call the politics of participation.  So who is invited to participate in those convenings?  Not only who has a seat at the table, but who is speaking from that seat at the table?  How much influence do they have on setting the agenda and the discourse of the discussion and the outcome documents?  Because it could be that the outcome documents and the discussion held at that forum could influence the real world, which also brings up another concern of many within these forums.  Well, what happens out of this?  It is great to come together and discuss these issues and it's critical to the process of defining what are the key issues and what viewpoints we should take, but we also need to consider, well, what is the what?  What are we doing here?  What are the outcomes?
    Launched in 2013, there is a strategy panel on multi-stakeholder innovation as ICANN goes through its transition to make it a more multi-stakeholder entity.  The goals of this multi-stakeholder strategy panel is to gather broad feedback on ways that ICANN can use tools and processes to make itself more legitimate.  What types of crowdsourcing or collaborative filtering or online collaboration tools could we use to make the process more equitable?  In another side of my research I studied how to build crowdsourcing systems for policymaking.  And, of course, all of you in this room know when we make an algorithmic choice, that is a choice and it can bias.  The way that we present ideas to participants for them to rate ideas in the crowdsourcing model, it could bias participants giving us certain types of ratings or response.  So how can we try to the best of our ability to kind of build out or design out bias?  Although I think it's a very difficult thing to build out.
    On the research question we are looking at how do online tools such as the IDEA platform challenge hierarchies and boundaries of community of practice in multi-stakeholder governance.  So the ICANN strategy panel actually deployed a crowdsourcing effort through the IDEA scale platform.  Have any of you heard of IDEA scale or used it?  By show of hands.  No one heard of IDEA scale?  Heard of it?  How many of you participated in the ICANN IDEA scale implementation?  Okay, I think that's a challenge that we have about inclusiveness of these systems.  So I will talk about that in a few slides.
    So when we are critiquing the design of these systems, we have to think about the structure of the system from a technical perspective, but also the social context in which that platform operates.  We decided to use Landemore's "Five Principles For Crowdsource Policymaking."  Those five principles are accountability.  Does the system communicate back to the users the insights being gleaned from the system as the process is going on?  You have people on there who are evaluating an idea.  Is it communicating back to the people the choices from the group?
    Inclusiveness.  Is the system easily usable by a large group of people, diverse people?  Of course, here in IGF can all stakeholders participate in an equivalent way on the system?  By default the system should be multilingual because our primary language is not just English.
    Is this system transparent?  Is it communicating to the participant how it is making decisions about the ideas that are presented toe participants for rating?  Or let's say, for example, there is some sort of rating schema where you can rate positive up, negative down, I do or do not support an idea.  Does it communicate to the participant that ranking?
    What I think is critical to these systems is modularity.  You shouldn't design a system of gathering feedback where you have to get on at the beginning and you can't enter ever again, right?  Because many of the participants may find out midway through about something and they've missed the opportunity to give their feedback.
    So the system needs to be broken up into a modular format so that participants when they find out about it, they can go into the system or into this collaborative process and contribute in a meaningful way.
    Synthesis.  In all crowdsourcing systems, the goal is to have an end product that comes out of that deliberative process.  So how are you communicating that synthesis back to the participants?
    Also should note that these five design principles do not just relate to the technical design of the system.  How it looks.  These are experienced online, but also about the process.  So are you communicating the modularity of the process to participants?  Are you communicating beyond having someone have to look through the system to see whether or not you say how your algorithm is prioritizing ideas?  Are you transparent about the process?
    Our initial rouse, we looked at five ICANN meetings, multi-stakeholder meetings from ICANN.  We also evaluated the final report and the supplemental materials of the strategy panel.  The blog posts that were published by the organisers, which were published on the Gov lab log.  And the IDEA scale platform itself.
    Here is a screen shot of the ICANN strategy panel multi-stakeholder innovation crowdsourcing.  So this is IDEA scale.  As you can see, there are ideas on here.  So participants were asked to put forward and idea for how ICANN could strengthen its inclusiveness and multi-stakeholder processes and how it could become more legitimate.  Participants were able to put in their idea.  You can see that other participants could go in and boost up the idea, vote it up or vote it down.  And they could also go in and actually dig in and comment on an idea.
    So the IDEA scale implementation had three stages.  In the first stage, it's this online IDEA stage scale system.  There were 80 participants, 20 ideas were submitted to the IDEA scale system, of which they received 46 comments and 72 votes.
    After this first stage, they went, the panel went forward with in depth public evaluation of those proposals.  Then they also did an internal evaluation and critique of those proposals, refined them, and gathered expert feedback.  Then they also had a collaborative drafting and editing of proposals on the gov lab log.  As you can see, this process was meant to be very inclusive and very transparent and very modular format.
    Quickly, I put this together to show the five design principles, and then evaluating this IDEA scale implementation on process and platform.
    So the process itself was very accountable.  There's a lot of reporting on the process itself.  However, on the platform it wasn't clear, on the IDEA scale platform it wasn't clear how ideas were first generated into the IDEA scale system.  So who decided to put those ideas in, was there a vetting process?  And then on inclusiveness, the process itself, as somebody who also does crowdsourcing in policymaking, it can be incredibly difficult to get a lot of participants.  Or you get one type of participant.  A group or a sector that is very interested in that specific issue.  It can be a challenge to open that up to make sure that you get the diverse perspectives that you need in order to have a more informed process.  Also the platform itself, the IDEA scale system was only in English, which does prohibit participants from entering their suggestions into the system.
    Overall, the process was incredibly transparent which is lucky for me as a researcher.  Everything is written up that I would need to do my research.  But the platform itself doesn't really discuss how the voting mechanism happens.  There's an up vote and a down vote and some comments.  There's actually a leader board.  So as I said before, there's this idea of politics of participation.  So when you have somebody who has greater influence or higher up in the leader board, are they going to influence the discourse and you have a winner takes all which is the worst thing to have in a system.
    Everything was modular and broken up.  The synthesis of how the ideas came, the final proposals -- it was very clear in the platform and it also showed this.
    So we believe these design principles are very useful for evaluating the process and the design.  We are looking now to extend this for our paper and really come with more of a sociotechnical lens to look at, yes, let's evaluate the platform.  Let's also evaluate the process and let's contextualise this within Internet Governance processes.
    So we are going to be doing interviews with the organisers of the ICANN IDEA scale surveys to evaluate perceptualisation of multi-stakeholder.  That's it.
    (Applause.)
    >> ALEJANDRO PISANTY:  I am apologizing for having started late in my participation in this session.  I think we should open for questions.  How much time do we have?
    We have 45 minutes.  Maybe some people will be glad if we make it 30.  If the discussion is ripe, we will go on.  I don't know, yes?  We believe in remote participation and using the Internet to make this meeting actually really open and diminish the barriers of cost and difficulty of travel.  So we will be very emphatic and listening carefully to questions from the remote or distance participants first.  And, of course, since some of them take a little bit of time to set up in a way that they can speak, we will be making several rounds of asking for remote participation.
    Again, I think there is a remote moderator present?  You have something?
    The remote questions are being sorted by the remote moderator.  We will start with a question or comment from inside the room.  I think we should first start with pointed questions and thus we see the time allows for more participation, we will allow for longer participations in the discussions, if all panelists agree as well.
    First, questions here locally in the room.
    If your beliefs systems allow it, it is very glad --
    (Announcement outside.)
    >> ALEJANDRO PISANTY:  We are waiting for some sound outside.  That is not us.
    If your belief system allows, would you introduce yourself?  It facilitates us.  I understand that some people want to remain anonymous.
    >> AUDIENCE:  Hans Klein from the Georgia Institute of Technology, School of Public Polling.  I have a question about algorithmic transparency.  The U.S. Trade Commission held hearings a year or two ago.  They considered an antitrust action against Google and they didn't frame it in terms of algorithmic transparency, but essentially it dealt with some of the issues you were talking about as well, which is the ability to, or the search results and using the search results to screen out competitors.
    So I wanted to ask you, has that issue come up?  I don't think you mentioned an antitrust perspective on algorithmic transparency, but it did seem to figure prominently in the U.S. policy arena.
    >> RACHEL POLLACK ICHOU:  So actually, the paper ... hmm, is that the mic?
    >> ALEJANDRO PISANTY:  For the sake of remote participants it is very desirable that you identify yourself every time you take the microphone, please.
    >> RACHEL POLLACK:  This is Rachel Pollack.  I presented the first paper on algorithmic transparency and search.  So yes, the antitrust perspective I alluded to a little bit.  I know, out of time.
    In terms of the European context with the competition case and the views of some of Google's, what I called industry competitors in vertical search.  In the U.S., indeed the Federal Trade Commission did consider a case against Google.  Ultimately it decided against pursuing it.  But the Wall Street Journal reported they received some of the -- they found that Google's contact has resulted and will result in real harm to consumers and harm in the advertising market.  That is something clearly that has been a focus for the moment.  It is only being pursued in Europe, but based on reports of the FTC and the FCC, I think again algorithmic transparency is related in the sense of being one tool that could ensure the neutrality or fair business practices of Google services.  But I don't think that that is ultimately what they would require.  They've also worked on, I think -- let's see.  They have an office of technology research and investigation that in March, launched in March 2015 that started to look at algorithmic transparency as one area of work.  From what I've seen so far, they clarified that it doesn't mean in the area of publishing of the algorithm.  But it is an area of interest.
    >> ALEJANDRO PISANTY:  Thank you, Rachel.  There was another speaker besides Hans?
    >> AUDIENCE:  Nick Suzor from KT School of Law.  My question is for Isadora.  Thank you for the presentation.  I enjoyed all of the presentations.
    Towards the end of your paper you touch on some of the normative implications a little bit.  You spoke a little bit about this.  I hope you can expand a bit on this point that you argue that the construction of negative liberty can ultimately be harmful to more positive steps by nation states to protect freedoms.  I'm wondering if for a moment you can reflect a little bit more and expand on your point there.  Thanks.
    >> ISADORA HELLEGREN:  So yes, can you hear me?  Yes, all right.
    I'm Isadora that presented the crypto-discourse presentation.  Thank you so much for your question.  Much as I tried to elaborate in my actual thesis, it is hard to fit it all into a paper and in a presentation.  But I do believe that there are many different variations of our understanding of freedom.  And I am from Scandinavia myself, I am from Sweden.  I believe the understanding of freedom there is very different from this more negative understanding of freedom in other countries.  I believe there are many different variations of what role we think that the State should have.
    By uniting many different political differentiating political objectives left and right, that all go against the state, I believe that it removes this responsibility that the State should have perhaps in protecting these online rights by saying that well, we cannot trust you anyway.  So it is our responsibility to protect ourselves.
    Then there is nothing left that would hold the state accountable.  So that's why I mean that there are many different understandings of freedom.  By forwarding this discourse, by calling it all Internet freedom on an international level, it could be harmful for other attempts to construct alternative Internet freedoms.
    Does that answer your question?
    >> AUDIENCE:  Yes, it does.  But I would like to read the entire thesis.
    >> ISADORA HELLEGREN:  Okay, thank you.
    >> ALEJANDRO PISANTY:  Let's see if we have remote participation, if you have anything already.
    Remote moderator is telling us that the user that looked like was going to ask a question has logged out.  We will wait for that.  And we'll come back to the room, if there are other questions.
    Sorry, yes?  Hi.
    >> AUDIENCE:  Hello?  Yes?
    Hi, lovely panel.  Very informative, all of you.  Marianne Franklin, GigaNet.  I was wondering, Isadora, could you elaborate a little bit more about what you mean by empty container?  Empty signifier?  Because there is one school of thought that would argue that no signifier could be empty.  By virtue of being empty, it is signifying something.
    >> ISADORA HELLEGREN:  What I mean by empty signifier, it is more of a process than a State.  So any signifier holding several meanings or so, going from a particular meaning, it is the process of emptying this particular meaning by filling it with so many meanings that the specific meaning becomes meaningless.  It becomes so big that it manages to encompass so many different perspectives, so many different approaches.  That is why I mean that the empty signifier crypto can encompass left and right.  There are crypto-anarchists, there are pacifists that believe or stand behind crypto, although they do not have the same political objectives, they unite in this particular objective of the social antagonism.  So the empty signifier is really where the struggle takes place.  Therefore, it becomes empty because there is so much meaning being fill in it.
    Does that make sense?  It is the process of emptying it and filling it.
    No?
    >> MARIANNE FRANKLIN:  Yes, yes.
    >> ISADORA HELLEGREN:  You seemed hesitant.
    >> MARIANNE FRANKLIN:  Yes, I understand what you mean entirely.  It is just that when we say empty, we imply devoid of meaning.  In fact your argument is saying it is over-determined and over-abundance of competing meanings.  The term empty in that sense you mean ironically rather than literally.
    I'm just fascinated with, perhaps you want to elaborate that conception more.
    >> ISADORA HELLEGREN:  Right.  Right, meaning that it does not mean anything in the end, but there are all these means.  I'm sorry.
    >> MARIANNE FRANKLIN:  Yes, this is great to have at GigaNet, because meaning in fact is the politics that we are dealing with.  Could you perhaps consider that the meanings that are there are competing and contradictory meanings?
    >> ISADORA HELLEGREN:  Yes.
    >> MARIANNE FRANKLIN:  Therefore, the whole issue is this competition and contestation rather than there is only one meaning, which is perhaps the terminology empty is not quite -- I know it is coming from other discussions, but meanings are always contested, all signifiers are containers for something.  What is interesting in your argument, and I take your point, is that it is exactly this over abundance of competing understandings of freedom that we need to be unpacking rather than seeing that in itself as a problem.  The real problem is that there is an assumption that there is only one meaning of Internet freedom.
    >> ISADORA HELLEGREN:  Yes.
    >> MARIANNE FRANKLIN:  And the struggle is to assert that one meaning overall others.  So the container is in fact overflowing with many competing meanings.  That's how I understand your work.  But thank you very much.  That was very useful.
    >> ALEJANDRO PISANTY:  I will use my moderator's privilege.  If it is not taken away from me from my dead, cold, fingers.
    I mean to mediate in this very important discussion between Marianne and Isadora.  From a much more pedestrian point of view I will say if you can't know what it means, it has become meaningless.  Because it is meaningless in itself because it doesn't parse into the English language or because you have to ask for so many more specifications.  Like what you mean freedom, like in the First Amendment of the United States, before the church, before the freedom of speech, so forth.  I would ask for other participations and I would have a question for each of the participants of the Panelists.
    So are there other participations?  First, again, to the remote moderator, do you have any questions there?
    Here in the room, do you have questions?  Then I will make them serial in one set an hope that sparks the conversation as well.  If you would, Daniel?  First, to Rachel, I find that all three papers, by the way, extremely interesting, very valuable.  I come from the technical community side but I'm sort of a bridge person in many ways.
    So one question since you have held so many interviews, maybe it could be a salient point even if you haven't, being quantitative about it, but do people who have experience coding react differently as a group to the concept of algorithmic transparency?  Do you see more coders saying that's a pipe dream, not achievable, not a desirable goal?  Not a useful thing to do as a class?  Let's say as a group?  Have you at least seen a hint of that in your data?
    Two, Isadora, there's a deeper comment that is sparked very much by Marianne's question, which is, when we analyze spam, one of the things we say is that one man's spam is another man's caviar.  What some persons find worthless message, other people will actually kill for to get it.  That's one of the difficulties with spam, by the way.  You cannot filter in any absolute sense and call something undesirable.
    I am going to go out on a limb here and I'm going to say that one man's rights is another -- or one person's, or one human's right is another human's threat.  So in some conditions, the concept of freedom from one culture may be seen as deeply threatening by another.  Do you observe some of that when you look at the crypto-discourse?  People are actually asking for the State.  You mentioned this very carefully.  You use words which in my very naive approach to political science are social contract wording.  And you are speaking about different social contracts with different responsibilities of the State.
    Does this very pedestrian way of describing it fit your paper?  Could you elaborate a little bit?
    And to Brandie, I was close to the process, I have already left the ICANN board by the time this crowdsourcing -- sorry, this strategy panels were set up.  I was a participant.  I was a member of the strategy panel on the role of ICANN in the intergovernance ecosystem.  I followed the other panels somewhat closely.  I found it striking there was very little reaction to the innovative approaches, strategy panel which was led by Beth Novac, which is the one you described, the word you described stems from.
    Do you see anything like a generation gap, sort of culture gap?  People that are more hipster-ish, more connected to technology, younger, react more and participate more in processes like IDEA?  Or is it just a language issue like you also mentioned, being in English, it could just not attract enough foam who -- people who don't speak English?
    And my second comment to you asking for your reaction is that the, as you mentioned in your introduction, the IGF process or let's say the process is called IGF in countries and regions are more executive and decision-oriented.  And do you feel that it is a threat for the concept of open deliberation that we insist on keeping in the global IGF?  Or can we manage and articulate both kinds of orientations?  Thank you.
    >> RACHEL POLLACK:  Thank you for the question.  This is Rachel Pollack, for the record.
    So I think that you raise a good point.  Those with more technical knowledge or experience tend to be a bit more skeptical about the idea of algorithmic transparency.  There was a paper written by some computer scientists together with legal scholars in the U.S. that argued that algorithmic transparency is neither necessary nor sufficient to ensure fairness because there are alternative techniques that can be used, technical techniques.  Also because of the complexity of the code.  Particularly when these algorithms are constantly changing with new inputs and personlised for users.  That would be technically very difficult to implement.
    I think it is not a universal consensus.  There are some computer scientists who advocate for algorithmic transparency and even offer to offer their technical skills in order to build open source, open algorithms and open search engines.
    So I think there is not consensus, but in general it is true that those with more technical expertise are a bit more skeptical.
    >> ISADORA HELLEGREN:  So first of all, everyone unites behind the idea of freedom.  Everyone wants freedom.  It is hard to oppose the idea of freedom especially when we are talking about liberal democratic states as we know them.
    What I talk about mostly are the strategies used to related to this idea of freedom.  In the negative, in the concept of negative freedom, strategies to enhance, promote, popularize the responsibility of the individual protection as an individual responsibility, as I said, removes any preference to the State.  If we look at Iceland, for example, and the international modern media initiative created by Pirate Party representatives, they use a different approach.  They still belong to the same, we could call crypto-community in many ways.  These people know each other, work together and do not necessarily have opposing views, although they might in certain issues as well.  It's a different context, a different form of freedom.  The international modern media initiative tried to protect the ideas of freedom of expression and privacy by reforming laws, using laws from other countries that are especially strong in regards to these questions.  It is a different strategy.
    So it is really the strategies I have been looking at when it comes to this.  That's in terms of your question, if there is a culture perhaps of feeling threatened by other ideas of freedom in other cultures, I'm not sure about the threatening part, but I believe there are strategies we can use to try to enhance our freedoms.
    >> BRANDIE NONNECKE:  Thank you.  I'm Brandie Nonnecke from UC Berkeley.  Thank you for your questions.  The first question was the limited participation.  I think one of the main issues for most of us, the way that we operate, is that when we want to get things done we go meet with the person face-to-face.  So we have these ingrained behaviors of how we believe we can influence.  So I think changing that idea of how you can influence organisations or have your voice heard -- I think that is definitely changing --
    (Announcements in the hallway.)
    >> ALEJANDRO PISANTY:  Sorry, the sound we heard is an announcement inviting people to go to the main session, the main room for the messages of welcome.
    >> BRANDIE NONNECKE:  Okay, thank you.  I think that the hesitation is valid because some of these tools could do more harm than good if those tools end up influencing actual processes and outcomes.
    So I think it is a careful balance between reliance on interaction in traditional ways.  So through our normal institutions of moving policy forward and even discourse.  But also trying to build out the tools in a way that mitigate bias and are more equitable.  I think this is going to transition over time.
    The second question was on what are the risks of having an action-oriented Internet Governance forum like I witnessed in the East African community.  Does this pose a threat to open deliberation?  I think in part it does because that forum has now become almost a secondary channel for stakeholders that are usually engaging with each other.  So government and private sector.  This is yet another channel for them to interact with each other and set policy initiatives.  But I think calling it out and knowing that it is happening is critical in order for other stakeholders to be able to go in and have an influence on that discourse and the outcome documents.
    I would also like to see not only the face-to-face interactions, engaging all stakeholders, but seeing how week use these crowdsourcing tools to help level the playing field for all stakeholders.
    >> ALEJANDRO PISANTY:  Thank you very much.  Are there further questions or discussion points?
    From the remote, anything?  We don't have anything from the remote?  We don't seem to have any more questions locally.  I think that both the presentations and the discussion after that have been enriched very much the announced content.  I don't know, Daniel, if we should just call the session closed?
    >> BRANDIE NONNECKE:  I don't have a question.  I have a comment.  Typically when I am at conferences or meetings, I'm often the only woman in the room.  I'm happy to see that in the room and also on my panel are other women.  I'm very happy for that, thank you.
    >> ALEJANDRO PISANTY:  Thank you very much.  This is indeed a trend that has been growing in the IGF.  The IGF has usually had, say coming from technical conferences you would say atypically high participation by women.  Of course, we have a ways to go.
    So if everybody else allows us, we will call this session closed.  We thank the speakers both for their papers and for their responses and replies to the questions and comments.  And we wish everybody a very good day and good stay in Mexico.  Thank you.  Any message from the Chair?
    >> DANIEL OPPERMANN:  Thank you very much, Alejandro.  Daniel Oppermann, for the record.  Thank you very much, Alejandro, and thank you very much for all the presentations and the questions.  We are going to close this session now.  We are going to have a 15-minute break.  After that we continue with the second panel on Legal Challenges.
    And there is still room here on the table.  People sitting in the back who would like to sit at the table, please feel free to join the table.  Thank you.
    (Fifteen-minute break.)

     >> THIAGO TAVARES:  Good morning, everyone.  So we restart the next panel regarding the Legal Challenges.  In this panel it is connected with the first one.  We have four presenters and we will follow the list.

     So I am glad to moderate this panel.  I would like to thank the organisers and especially my dear friend Daniel Oppermann for the invitation.  Well, the idea is to have one short presentation, about ten minutes, for each paper.  Then we open the discussions for questions.  We have remote participation as well.

     Let's do it.

     Please, Markus, you have the floor.

     >> WOLFGANG SCHULZ:  So thanks very much for the invitation.  I hope it is -- hmm?

     Thanks so much.  Seems to be better this way.  Thank you, Chair, thanks for the invitation.  What we want to do is to present a project outline doing governance in configurations, constructive normative structures inside and outside of intermediary organisations.

     This project does not stand along.  It's part of a configuration.  A bunch of researchers who want to study social domains and social reality with the same theoretical background and same analytical framework.  Our aim is to better understand how governance structures or normative structures evolve and want to apply the concept that Markus will explain in a couple of minutes as well.

     Our starting point is that we try and analyze phenomenons like the normative structure that has evolved after the right to be forgotten discussion by the European Court of Justice that has dramatically changed the way in which at least specific norms have been constructed.  We now see that inside of the intermediary organisations involved, norms evolved that have been part of other spheres that have been constructed with the help of other actors.  We are interested in how that works, how that interplays with the normative structures outside of the organisations.  And this case, this search engine case is a very good case study for us because here the difference between constructing norms inside organisations and outside organisations and the interplay between both is very specific and you find very few examples of these kind of norm production elsewhere.

     How we want to do, Markus will explain.

     >> MARKUS OERMANN:  So there was a little bit of confusion.  I am Markus Oermann.  For everybody who is reading, there was a twist in the name.  To the concept.  Our main research question is how the normative structures are constructed.  We take social constructivist starting point and we ask ourselves, okay, how could we do this governance research analysis based on that what we already found in the governance literature?  We see there that governance is already developed as a counter concept to hierarchical control and taking governance factors of different kinds into account.  We differentiate between laws, contracts, social norms and technological code.  All of these factors together are combined in governance models today.

     So this is what we already have in governance research, names change from time to time, but yeah, we have to structure the view.

     Also we have the view on multi-stakeholder formations and so on.  But there are some shortcomings still left in our point of view.  From our point of view governance research today is still biased towards understanding structures without taking the process perspective into account.  Research on organisational governance is focused on structural changes and their management inside organisations.  So business organisations, so to say.  These studies ignore the environment of the organisations from most of the times and from our point of view research on Internet Governance especially is focused on multi-stakeholder governance structures, entries, organisations as monolithic structures and doesn't have the inside view on organisations, especially on business organisations.

     So we try to overcome those shortcomings by developing a figurative take on governance.  This concept is based on process developed by nor Bert Allies and developed in the last years by others.  And they say, okay, the governance structures are constructed and they are constructed by actors.  These actors form together figurations.  We now try to look on the processes going on in this actor figurations.  They are characterized by three main criteria.  So it is the actor consolation, the communicative practices going on between the actors.  So our standpoint is that communication is the main mode, how governance structures are constructed.  And on the frame of relevance, so to say.  The thematic framework that binds these actor together, we have actors like the European Court of Justice, Google, Civil Rights organisations and they construct the normative structure that is relevant when it comes right to the thematic frame of the right to be forgotten, so to say.

     That's our mainframe work.  This provides us two method logical points of entrance to the field.  So we can analyze the materialisations of the factors, the laws, the contracts, the clauses, there is close connection to the paper of Nick in this point.  We can do process analysis.  This process analysis could be done by interviews and also could be done by field work going to the organisations and looking at the actual process going on inside this organisation, like the European Court of Justice and in our research, inside the business organisations.

     So Wolfgang will expand on that point now a little bit further.

     >> WOLFGANG SCHULZ:  Okay, it's me, Wolfgang again.  Now, coming back to the case of the right to be forgotten, as you could see from Markus' talk we have two structures that we are looking at.  One is the traditional base for factor model of Internet Governance and we apply that as well.  What we add to that, we look for each factor and the interaction of the factors to our figurational three-level model.  We analyze empirically what actors are involved inside of the organisation and outside of the organisation, and how they interact, how they communicate, what communication media they use.

     The second thing is that we want to analyze what the analytical frame actually is under which they operate.  Do they operate under the assumption that we have to create a right to be forgotten?  That would be the frame of reference that is relevant to us.  We would see that the norm making structure under this umbrella and the last thing is that we want to identify specific practices of norm making.  We want, for example, to see how inside the search engine organisation these norms are formed.  We know that it will not be easy empirically.  Our dream is to have someone inside, ethnographically working on that and seeing how the conflict is processed when there is a new case coming in and how the new norms to decide on these kind of conflicts evolving in the organisation.

     So that is basically what we want to do.  If everything works out, then in this group of researchers we have the chance to compare that to other norm making procedures.  For example, there is one colleague working on religious authority and the construction of religious authority.  I don't say that intermediaries are churches, but there might be some similarities as regards norm making in different social domains.

     Thanks very much.

     >> PATRICIA VARGAS-LEON:  Good morning, everyone.  I am Patricia.  I come from Syracuse University.  Thank you for the opportunity to present my work-related to the implications of the application of the hot pursuit principle, a mechanism of the governance of the cyberspace.  Everybody can hear me?

     Is it better now?  Can everybody hear me better?  Okay.

     Okay.  So I will start my presentation by acknowledging a fact that probably everybody hear knows.  Which is that in today's world there is no treaty that regulates the Internet or the cyber space.  Although the multi-stakeholder model has been successful in keeping the Internet free of a unique point of control, there are still nation states and individuals who are looking for a government-based model.  This scenario was clear during the last ICANN decision debate when some nation states argued in favor of IGF being managed by the ITU.  The first question I came to in analyzing this, why these different opinions exist?  I believe it is because nation states have different conceptions of what should be the governance model for our resource beyond the traditional border of their own territories.

     In concrete when it is about cyber space, academics and nonacademics, mostly militarily, have called to apply similar policies to the ones applied in the UN Convention on the Law of the Sea, also known as UNCLOS, into the cyber space.

     What reasons are they giving is the next question that I asked myself.  Well, first of all, they argue that the sea was once an unregulated space beyond nation states' traditional borders.  The sea is also a space for war, economic production, and communication.  And it is constantly subject to sovereignty claims.  Pretty much as the cyberspace today.

     As an example, in June 2016, that is pretty much six months ago, NATO announced that 28-member alliance agreed to declare cyber an operational domain, just like the air, the sea, and the land.  In the United States, you mind very often the military referring as the cyber power, the cyberspace as the fifth element.  They talk about sea power to cyber power.

     But the question now is what is behind the potential regulation-based in UNCLOS?  Why is it so constantly defended by nation states so different such as well established democracies, hybrid or authoritarian regimes?  The first thing to say about UNCLOS is that it is a multilateral treaty, a UN treaty that covers multiple aspects of the regulations of the spaces and activities in the ocean.  I am going to ask you to please look at the map -- I'm sorry?  Okay, to look at the map that you see up in the Power Point.  I know it is very confusing, but that is the distribution created by UNCLOS.  The main idea, the major characteristic is the creation of fictional spaces, where nation states keep specific sovereignty rights for them.  The idea is that coastal states sovereignty decreases with increasing distance of the coast.  The end of this sovereignty claim is either the high seas, an area that belongs to everybody, or another state's territorial waters.  This policy is known as maritime jurisdiction.  I, on the other side, call it fragmentation of the ocean, taking a new term by the academics, we can call it a virtual alignment of the on the grounds.

     The hot pursuit principle, a mechanism to fight against illegal activities in the sea, is an element of UNCLOS and was built on the basis of absolute respect of nation states' sea sovereignty rights.  So trying to make a similar or potential comparison, I argue that the potential implications of applying UNCLOS into the cyberspace include, number one, a call for the inclusion of nation states sovereignty and jurisdiction.  In other words, bids are under nation state jurisdiction.  When controversies arise, leading experts -- we can see a very clear case, as I argue, in the dispute between Microsoft and the United States, also known as the Irish case.

     We know that the Internet is invisible and, the application of UNCLOS may have implications that manifest in potential fragmentation of the Internet infrastructure.

     And finally, I think this is from my point of view the most important and dangerous implication, is that by having solo actors to nation states, the application of UNCLOS into the cyberspace has as a consequence the creation of a potential model opposed to the multi-stakeholder model, that for good or for bad, criticizes as much as it can be, is the one that has allowed us to enjoy a free and open Internet as we have begun to know.  Thank you.

     >> NICOLAS SUZOR:  Thank you very much, everyone.  I'm getting a signal to move across.  I can't see most of the cameras.  I'm Nick Suzor, Associate Professor of Queensland University school of law.  Do you mind if we squish over a little bit more?

     Can I move this?  Thank you.  We can change seats.

     How is that?  All right, thank you.  Thanks for bearing with me.  Okay.  So my name is Nick Suzor and my project is about legitimacy, specifically my conceptualisation of what I call digital constitutionalism.  It is used in different contexts by different people.

     But the core point I want to make today is that we need better theory to start to understand what legitimacy means in these private governance systems.  This is actually a conversation that is really difficult to have.  It is a conversation that is quite topical at the moment particularly when you look at the increasing power of private online intermediaries, platforms in governing the behavior of users.  And so what I think is missing fundamentally from this discourse is a way to articulate what legitimacy means and what sort of standards of legitimacy we might expect from private actors.  The key here is, I don't want to suggest that the standards to which private actors ought to be held is the same standard that we hold nation states in public systems of governance.  But I do think that those values help us to particularize the issues that arise so that we can start to add some rigor and some clarity to the discussions about the things we care about when governance is done online by private actors.

     So that's the general gist.  The first point I guess is probably to clarify what I mean when we talk about governance.  By governance, I mean to understand it in a very broad sense.  In the sense of the management of the course of events in a social system.

     So the task there for regulatory scholars is if we recognize the governance is decentralized, to then come up with a set of indicators, and a theory, a framework to understand what legitimacy might mean in a decentralized system.  And legitimacy also has multiple meanings.  One component of legitimacy is a descriptive component.  The people who are doing the governing are accepted as having a right to govern in the way that this he do.  I'm more concerned here with the normative components of legitimacy, the set of contested political standards against which we judge the exercise of power.

     So this then is my project of digital constitutionalism, which is particularly to start to articulate what limits on the exercise of private power we might like to see in decentralized systems.  This is about the formation of a polity and the rules for how governance is done, constitutionalism here in the sense of a group of people who are bound together and thinking about the constitution, the way in which those groups are governed and organised and set up.

     So the short version of the paper, the rules of the values of rule of law, the procedural values of the rule of law can help to articulate these questions.  The rule of law is the concept that those who exercise power over our lives, governance power must exercise it in a way that is equal, certain, and fair.  That is the core liberal set of values of the rule of law that are often treated, primarily by liberal scholars, as universal.

     But that universalisation has a certain claim.  The a least in some concepts, some concepts of consent of the governed, I guess, is probably universal in systems of governance, as a core standard of legitimacy.  So in the paper I present this framework.  I'm happy to talk about more in questions or after this session, but basically articulating three different buckets of concerns of things that we might care about in terms of procedural legitimacy.  The first is meaningful consent.  That rules are clear.  That there is genuine consultation on changes.  That there are real opportunities to exit systems.

     The second is about equality and predictability.  The decisions are not arbitrary.  The rules are fairly and equally enforced and that rules are stable enough to guide behavior.

     The third is about due process, that there are reasons available upon which decisions are made, that participants have some access to some avenue of appeal and independent review.

     Again, to be clear, I don't think, and I don't want to suggest that we ought to hold platforms and businesses exercising governance power to the full high standards of legitimacy that we expect from nation states, but I do think that by particularising these values we can provide a language to start to articulate some of the tensions that we are seeing in a way that is a little bit clearer and a way that helps us advance a political process.  This is really a highly contested area.  We don't have agreements about the standards to which in my case platforms ought to be held.

     And it is not easy within this system of private governance to talk in a language of rights when rights are traditionally applied mainly against public state actors.

     So by using this language I hope to be able to add a, at least a shared understanding of what is at stake even if we disagree on the implications of that analysis.

     The work in the paper focuses mainly on the formal terms of service documents of social media platforms, looking at this stage just at the legal documents.

     On that level, the spoiler is that judged against these values, contractual terms of service are actually really poor constitutional documents.

     The next step, though, is really to start studying governance in practice.  I am happy to be on this panel.  The work done by Markus and Wolfgang in this paper I think is the key to how we need to start to understand what is going on.  One of the challenges, the big challenges in talking about legitimacy in governance in the space that you pointed out is that we don't actually know a whole lot about how private platforms make decisions.  Most of what we know is based on anecdotes and leaked information and a few investigative journalists who have been able to get access, and then the public documentation that these different private actors submit to public processes and make available online.

     Studying governance in practice, I think, is the next required step to be able to actually then articulate what we might believe to be a shared set of political standards for how we think that private actors ought to behave when they are governing the lives of participants.

     I could talk a little bit more about different methods, but I'll leave that to question time.  And I'm happy to take other questions after the last speaker.  Thank you.

     >> MICHAEL DICK:  I'll try that.  How is that?  Not loud enough?  We have to try the wireless one.  I don't think the cable will stretch.  It needs a longer cable.

     Or we'll just do that.  Okay, perfect.  Thanks very much for inviting me.  This was sort of a last minute invitation, but I'm very pleased I was able to make it.  My work online is an extended extract.  My name is Michael Dick.  This research I'm about to present in a very brief form relates to my dissertation on the modernization of Canada's copyright act and the various issues inherent in both public policy writ large and the policymaking process, consultations and so forth relevant to this particular exercise in policy reform.  I'm at the University of Toronto with background in communication studies, cultural studies and media studies.  I became interested in issues while researching technical standard setting.  I was interested then and am now on how this takes on a de facto form of gov and interacts with manifest tomorrows of Intellectual Property and especially in my case copyright, which has a lot in common.

     I want to suggest to you today as part of this panel that IP intellectual policy, Intellectual Property policy, IP policy and Internet Governance are actually interrelated issues.  I want to demonstrate by offering a few key insights into a case study pertaining to Canada's new copyright regime.  I appreciate this is a matter of more local concern.  You may wonder what it has to do with global Internet Governance that we are tackling at an event like this.  This brings me to a point that is obvious but I had to for the sake of academic rigor confirm through a decade's worth of policymaking materials that brought us to legislation known as Bill C11 in Canada.  Brought into law in 2012 and introduced several new economic rights into the copyright regime.

     And this important point that is brought into that is the fact that WIPO, the World Intellectual Property Organisation, actually played a very direct role in a national level policymaking and now governance process.  Much of the discourse on copyright reform that dominated the policy landscape in WIPO signatory countries over the past two decades has been shaped by treaties known as WIPO Internet treaties adopted in 1996 across Member States.  The U.S. adopted these quickly with the implementation of the Digital Millennium Act.  It would take Canada many years to do the same, due to a myriad of political issues, governments rising and falling, et cetera.

     What is interesting here is that the bills were basically adopted piecemeal at the insistence of industry lobby groups and consternation of copyright activists and user groups.  The federal government promised widespread consultation as part of the copyright initiative, the extent to which this was actually balanced remains the subject of debate.  My particular interest is in the implementation of a new copyright known as Making Available Right or MAR, essentially brought about to bring copyright into the Internet age by allowing rights-holder to restrict making content available streaming, some form of exploitation of that content over the server, per se.

     In Canada the industry lobbied for this right to protect against file sharing and what they viewed as piracy, thinking that people could be held for infringement for uploading content to a server regardless of what end users do with that content and whether or not that proved to be infringement.  That is a complex issue that needs to be addressed through the courts as we know.  Alternatively, rights advocates could monetize the act of making content available by going to the copyright board of Canada, our regulatory board, and pushing for certified tariff that would require uploaders of files to pay royalties for their actions of doing that.

     I am happy to talk more about the national level specifics separately, but given the nature of this forum I wanted to make this point, that a super national policymaking body effectively dictated the treaties, although I don't believe this' what they set out to do.  To be clear, my criticism is with the Canadian modernization process and not WIPO.  I understand how rights like the MAR found themselves into the treaties, which is a subject worthy of discussion.  Why did Canada do what it did?  Due to the result of lobbying efforts of industry but due to the perception it had to meet the requirements of other groups regarding Intellectual Property.  The existing right of communication to the policy within the IP regime covered infringement of usage.  This is an important distinction to make, that there were indeed other options.

     Now that we have this in mind I want to make my final point today to bring this into the space of Internet Governance and the legal challenges inherent within it.  This is approve indication as well that I hope we can unpack further in this panel around the point of looking at this in practice because it gets to the heart of technology neutrality, which I dare say is something that concerns all of us here both on the panel and in the room.  I suggest in my work, I think above all else, the implementation of new rights like the MAR without fully understanding or studying implications, to do it because WIPO says so for lack of a better way to put it, is problematic.  It is not just WIPO, bodies like IHAC, the Information Highway Advisory Council in Canada are recommending changes to the IP regime that protect the rights of content creators in the digital age.  They have good intentions, and while protecting the rights of content creators is important, but it doesn't reflect the reality of today's climate in the Canadian landscape.

     I develop in my work the argument that this takes on a form of de facto Internet regulation and in spite of the fact that Canadian broadcast and regulators have opted for positions of forbearance on digital networks, and they have restated the point several times.

     Content regulating the Internet?  Really?  Not so far fetched when you consider this has been addressed through, for example, Lessig's notice of legal liability and take-down notices, but this is a case where right implemented in a country's civil law framework adopts a guilty until proven innocent through online networks by way of server.  Similar phenomena in the analog space or offline spaces have been litigated.  Key Supreme Court decisions held that making content available or the means of making that content available is not prima facie infringement of laws.  One of the colleges that regulates lawyers in Canada was not liable to a group of publishers for quote-unquote copyright infringement by making research services to members.  Here their act of placing a photocopier in the library, making means available to infringe the copyright, was seen as fair dealing, or what other countries call fair use, a substantive defense against copyright infringement.  Let's bring this to the online space.  If the means to make content available is not infringement, what is the merit in regulating the content online through rights like the MAR?  How does this level of restriction conflict with the open Internet and web that we want?  How does it conflict with the notion of neutrality and rights should remain focused on the use of the object rather than controlling access to it?  These are main questions we should address.  They should have been taken up before Canada adopted the WIPO treaties wholesale at the interests of groups trying to use the MAR as insurance policy through which online file share could be controlled even in cases where the courts and other bodies found that certain ways in which content is communicate the to the public via Internet enabled or facilitated means were not necessarily controllable nor compensable.

     Most importantly we must consider the elephant in the room.  This is a form of policy regulation or whatever we want to call it that was partially devised because copyright regimes could not keep up with the means that we have to reproduce share, stream, and otherwise distribute content in the rapid new ways that the online environment affords.  So will this change with new rights like the MAR?  I think not.  I think the chilling effect this could have on the balanced approaches to online governance and copyright that assuredly most of us in the room ultimately want makes them dangerous.

     I look forward in that spirit to continuing this conversation with you today.  Thank you.

     >> THIAGO TAVARES:  Thank you very much for your presentations.  It was great.  I have some questions, but I will keep it for the second round.

     And I would like to open the floor for comments, questions.  We also have remote participation.  Maybe we have some questions from the remote participants?

     Yes, okay.

     So the floor is open.  Please, raise your hand up.  Yes.

     >> AUDIENCE:  Hi.  This is Rachel Pollack.  Thank you for all the presenters, very interesting research.  I had a question regarding the first presentation about the normative frameworks within intermediaries.  I don't now remember if you mentioned it or I saw in your paper that you intend to do interviews and also ethnographic research.  Have you already started that process?  Has Google been open to accepting an ethnographic researcher?  My experience from what I reaped and also personal experience is that they can be quite closed about their internal processes.  I would be curious about how you intend to tackle that.

     >> Yes, quite right.  It is a mixture of methods.  We have some three levels, three phases of interviews.  We also want to do as icing on the cake these ethnographic things.  We have not settled that so far, but we had some informal talks so far.  I think it might be in the interests of the organisation to become a little more transparent as regards that.

     But the project is designed in a way that we get some meaningful results even if we cannot have these ethnographic approach.  It would be great.  They are, under the German legal education, for example, you have these so-called (German word.) and Google takes those people on board.  There may be one approach to say that there is one of those scientific junior researchers who have their reference there and make use of that for their ethnographic research.

     But of course, you hit a very important point there, but we are quite sure it is possible to get some valid information without it.

     May I make a comment to this question?  What we found really interesting when thinking about and trying to organise this project is how the organisations are dealing with their boundary work that is going on.  Do they close-up or do they open for the public?  And especially the right to be forgotten is an interesting case.  So they found it more or less public extra organisational expert comment on the right to be forgotten in Europe.  From our point of view, this is also a process that could be seen as the organisation opens for the public because the normative structures transfer this power to inside the organisation.  So the organisation is forced to develop norm sets.  These norm sets are related to public interests.  And this is forcing the organisation to open for public discussion.  And we hope that our project could work on that too.

     >> NICOLAS SUZOR:  I want to add to this how much of a major challenge this is right now.  For something that a lot of people think is quite important, it is incredibly difficult to get good information about how private decisions are made.  We have been supervised a few Ph.D. projects and trying to get access for many years.  We were unable to really get access to the major companies.  We were better able to get access to some of the smaller groups.

     Worldwide only a handful of researchers have gotten real good qualitative data out of large-scale intermediaries about decision making processes.  I would love to hear how you go with that.

     Then for me, as a question of research design, there are implications from this as well.  To an extent we have given up on some of the internal qualitative approaches because it has been so difficult to carry out that we've resorted now to try to develop new black box testing methods to watch from the public outputs of governance systems, thinking about what material is removed to be able to develop monitoring systems, I guess.

     So particularly for content moderation or for DMCA take-downs, for example, where we are unable to get that information directly from providers, there are opportunities for use of digital methods to try to reverse the internal processes.

     I probably should also mention that some intermediaries are much better at providing this information than others.  You have to acknowledge that Google and Twitter, particularly in their transparency data are quite good at reporting at least on external requests by governments.  Part of that is a strategy of resistance, particularly to European courts, I think.  But it does provide a lot of information for researchers that is not available for some of the other platforms.

     So I think this politics of transparency is also really interesting as it starts to evolve and people need to adjust.  Increasingly, in order to maintain that legitimacy I spoke about, the belief amongst participants that the platform is governing in a way that is legitimate, increasingly we'll see more disclosures about that which will be useful for us as researchers to be able to dig into what is going on.  It's a tough methodological challenge.

     >> AUDIENCE:  While we are on methodology, access always has to be negotiated and maintained, particularly for private situations. If you were to gain access in some respect, what would it actually involve, this ethnographic approach?  Speaking as someone who has done extensive ethnographic work?  If it is possible, what would you be doing and over how much time and which degree of immersion, if it were possible?

     Perhaps if you had any comments on that too.

     >> MARKUS OERMANN:  My staff was referring to Nick for one or two sentences.  In this case, it might be really good test case for Google here because as far as I hear from the data protection offices in Germany, they are quite content with what they are doing.  So the norms they are evolving there more or less are in line with the data protection regulation as the data protection officer sees them.  It is not a kind of investigative research to undig something which might be harmful for the organisation, but it is more to really understand how the norms are developed.

     So it might be different to other internal organisations you have with a company.  But coming back to your of course very important question, we thought about that a lot.  We want to build on the research that our colleague did.  He is the only one who already did ethnographic studies on intermediaries, on online intermediaries.  So we want to build on what he did.  We are not sure that we can do it as long as he did it.  I am not sure, it was several months, nearly a year I think he was in this one organisation in Great Britain, but the minimum we want to have is about four months and more would be better, of course.  And the focus is we have not let out in detail, we must confess.  It is in a state that we have this research outline.  What we really want to understand is to map the cases when the new decision has to be taken and there is a discussion, internal discussion about how the rules that are in place have to be applied or modified.  And when this modification takes place, then we find it extremely interesting to see what are the intellectual sources, the normative sources that are used.  We have heard already that German role plays a law.  A lot of cases come from Germany.

     Of course, the American tradition plays a role.  There might be some other bodies of rules that play a role.  We are especially interested in, to see how a new set of rules emerges inside and organisation, making use of external normative sets of rules.  And that is the main focus of this research.  That is what we try to understand.

     >> NICOLAS SUZOR:  I would say I am not an ethnographer, there are a lot in the room that have done work to understand these practices.

     My research design is primarily around comparative analysis of qualitative case studies, particularly looking to understand the influence of different regulatory frameworks and different external stakeholders on internal governance processes.  So, for example, trying ton understand how a formalized notice and take-down scheme under copyright might lead to one type of decision making and slightly different scheme under the right to be forgotten might lead to a slightly different type of process for decision making.  And then comparing that with the unofficial soft law co-regulatory and simply pressure from civil society on intermediaries to govern their networks in particular ways under different bodies of law or sometimes not under legal regimes at all.  Thinking about hate speech and vilify indication, some of which have dealt with complete outside of legal frameworks, and trying to use that to compare across intermediaries and across different areas of regulation.

     >> AUDIENCE:  Thank you.  It's very interesting.  I'm being academic here about methodology.  All of you appear to be doing what some schools of thought would call archeological work.  You are reconstructing, drawing inferences from sometimes a decade of documentation, formal and informal.  This is an important point to make.  Now that these spaces have been around that many years, the official and unofficial record is now available.

     And in the case of Patricia, your work is perhaps flagging the next set of archives, formal and informal, that will be perhaps discovered and reconstructed.

     My question is once again, where do the human actors figure in this archeological approach to the document from a legal point of view analysis?  I'm not saying they should figure, but they seem to be implicitly part of your frameworks.  Do you see yourselves or have you done interviews to check?  Just if you could extrapolate a little bit to help all the students and researchers in the room.  Patricia, perhaps if you have anything to note on the methodology for your work because it's fascinating.  By the way, Marianne Franklin for the record.  I keep forgetting.  Very rude of me.

     >> PATRICIA VARGAS-LEON:  Thank you for that question.  Addressing the issue of the methodology which is the one you pointed out, thank you, women, this paper is legal work as you can see. The main subject of data that I use was the set of deliberations that which was UNCLOS.  This is as a result of 100 years of negotiations, three conventions by the UN, one by the League of Nations and 2000 years of state practice, because 2000 years is obviously too much.  And because the League of Nations no longer exists, I based my analysis on the deliberations of the last convention, from the three that the UN conducted, only the last one, which took about 20 years.  It is essentially an extensive work.

     What I basically did was taking what the literature review has mentioned so far about the application of UNCLOS.  In concrete, I think there are three points of analysis.  One argues about the application of UNCLOS to the Internet packets.  I believe that is going to create more confusion than helps, but it is still part of the academic work.  Somebody else, another academic argues about using the rules to fight against priors under UNCLOS into priors in the cyberspace and also the military.

     The three aspects of the literature has been covered so far.  I come up with implications of the application of the hot pursuit and concrete.  This is a mechanism to fight against illegal activities.  But the main characteristic of the hot pursuit is that no matter what happens, and this is the way the ICJ and the International Tribunal on the Law of the Sea acknowledge it as well, is no matter what the crime is, no matter what is going on, no matter even the gravity of the crime, once the sovereignty of the third nation state starts, the other has to stop.  That does not mean that one nation state cannot access the data in another nation state.  This is taking it back to the cyberspace.  But you have to request through the traditional mechanisms of the negotiations between nation states, right?  In the case, concrete case of the United States is the use of the mutual legal assistance, all the bureaucracy that is behind them.

     >> MICHAEL DICK:  Thanks very much for the question.  It's an excellent one.  I like the way you described that as archeological work.  I think that especially in my case, looking at the reform of an Intellectual Property regime of a copyright policymaking process, a lot of this material obviously is publicly available in the sense that it is part of a legislative process and part of a policy cycle, if you will.  In my case it was basically looking at and sort of mining, doing a deep dive, in some cases using coding structures, in some cases using more grounded theory oriented methods.  Looking at basically two main consultation phases that were run over the course of this approximately ten-year period.  And there was sort of a second consultation phase launched because, as I said, because of governments rising and falling over that time.

     And there's in that framework hundreds and hundreds of stakeholder submissions from various actors.  Actors are very important obviously.  We have been talking a lot about that in this panel.

     One of things I wanted to say as a comment on methods is that it is not that you can't take that material at its face value.  It is that I think we have to put this into the broader understanding of agenda setting and I think -- I really think this is the case on many topics, but in the case of copyright, having been involved in the process of preparing expert witness testimony in court, in copyright board meetings and being part of that process myself, decision makers are basically alive to whatever you ask them to be alive to.  And they are alive to whatever their broader agenda and their govern ors are send -- govern nor are sending down to them.  For every person arguing it should be this way with a copyright regime, there is somebody on the other side and saying something different.  For every Metallica wanting to sue Napster, there were musicians saying this is good for our living and our brand.  It is promoting us, free publicity.  You are always going to find actors with sort of object sit perspectives on many of these issues.

     I think the challenge from a methodological standpoint is trying to weigh that as more of an objective scholar or trying to be an objective scholar and trying to suss out the agenda or the motivation in that without making sort of drastically off-base conclusions.  This is, I think, the issue sometimes with doing things like interviews to triangulate that.  You still face the same agenda setting.  People who made decisions at the time because they were part of an industry lobby group, for example, in many cases they are going to back that up because they are still working in that field or nobody likes to basically say I was wrong, oops.

     Although, actually, just about a week ago -- no, a month or two ago Michael Geist, a Canadian individual, a lawyer, a copyright or copy left activist working in this space, posited that some of the folks who were lobbying as part of the music industry efforts in Canada for the new making available right and the wholesale adoption of the WIPO treaties, they were basically contradicting themselves when he analyzed and compared a couple of their speeches to various economic clubs.  He basically was trying to point out the fact that without actually admitting it they were all of a sudden saying oops, this maybe wasn't the best approach.  We should have thought something through a little bit more.

     You have to be a built of a sleuth and detective when you do some of this work sometimes because people will seldom come out and say well, on second thought, it doesn't happen that often in my experience.

     >> MARKUS OERMANN:  Just a short word how we identified the actors who are -- this is Markus speaking.  We do analysis as a starting point that covers the right to be forgotten decision.  We have relevant actors in this procedure going on there already.  And we approach the actors that are named in the media discourse as relevant in the first phase of guided interviews.  Then we ask for what other actors that you find also are relevant that we didn't cover in our first wave of interviews.  Then we do a second interview wave.  It is always a hard task to bridge the levels of individual collective actors and how they interact and how they reconstruct themselves and the other actors in the field in the discourse.  We have to take a starting point and ours is the media discourse and then the second analysis.

     >> NICOLAS SUZOR:  Thank you for the question.  There is also, you know, no simple answer to that which I think may have been part of your motivation in asking the question.  There are different approaches for different tasks that we particularly employ.  And I guess at some level some of the work I was talking about today, most recently we are an abstracting away from the individual humans in the system.  That is a choice that we have made in that research design.  It is a weakness in the ability to draw conclusions about what is going on inside the every day practice of governing a system.  So I really like to see the rich ethnographic work of people who are able to get in there and document that.

     We do see roles of particular actors come out from a lot of the literature, a lot of the gray literature work too, particularly as you see how much of the ideology, how much of the framing in which the way -- sorry, the way in which the platforms understand their role in governance and are willing to conceptualize their work and willing to engage with external stakeholders really changes with changes in CEOs, changes with other influential people within an organisation.  You do see that come out quite clearly sometimes from the textual analysis.

     Other times we are really interested in how people participate.  So thinking now about the subjects of governance here, how people participate within different governance regimes.  I like the frameworks like the work on participation.  We like to look at the day-to-day acts of governments in networks and we can collect traces of that, particularly digital traces for social media and difference different platforms to understand a little bit more about the contest particularly, about legitimacy in governance.  But it is a fascinating area, and there's many, many methods to choose from.

     >> AUDIENCE:  Hi.  Alison Gillwald from ICT Research in Africa.

     My question is for Nick, but it also relates to Markus' group, their collective paper.

     I think a lot of your discussions that are happening around legitimacy and making the government systems work better, et cetera, we continue to do, apply the same normative frame to it, which doesn't really address those who are feeling marginalised or exclude themselves from these processes because they don't accept those normative frames.  So I suppose there is a sense of which, the work that we are doing continues to perfect the system or improve the system for those inside but doesn't really engage with how we get those who don't support, for example, multi-stakeholder governance on board.

     So I think the idea of holding private companies as key stakeholders and decision makers accountable is potentially a way of bringing those people not necessarily who, no normative frames you can draw on at all.  The sort of administrative justice systems you expect, rule of law, et cetera, are not things you can appeal to.  People therefore don't want to be participating in the multi-stakeholder systems.

     But some sort of standing on the cusp of things and feel that states should be holders of this kind of governance, et cetera, that perhaps at least if you could be arguing that these private platforms, private companies, et cetera, have some obligation around transparency and accountability you might be able to draw in some people who are on the margins, large numbers of countries who don't see any legitimacy and credibility within the system.

     So the question then is, what is the mechanism to oblige public platform, private companies' platforms, et cetera, to make those disclosures, to be transparent in the sort of unregulated environment?

     And how could you link that possibly to sort of broader disclosure and sort of developmental issues.  For example, public data or big data, something like that.  So it is not something discretionary that is given out to developing world governments in order for certain platforms to operate.

     But there is a sort of quid pro quo as part of this transparency.  Is there some way of actually making it happen?

     >> Thanks.  It's such a terrific question, important one.

     >> NICOLAS SUZOR:  The short answer is I agree with you completely.  The first step is transparency.  That is where we as civil society, as people who are concerned about governance need to be applying the most pressure, I think.  Before we can even have any of these conversations about the massive contested political values that are at stake, the substantive rights and other values that are at stake here, we need to understand better what is happening.  Only then can we really have a more -- well, a more political discussion about what should we expect.

     The slightly longer answer, I guess, is one of the things that brought me to this framework in particular is that the idea of, I guess, the recognition that when we have some of these debates with people who are not necessarily in this room or in rooms like it, it is hard to have a shared language with which to discuss these issues.

     I'm working with colleagues, scholars with a lot of experience in domestic violence, for example.  Recently I think the dialogue between Internet Governance people, and particularly people from the tech sector, has become much better at engaging with the substantive concerns of victims of domestic violence or abuse.

     That has happened relatively recently and for a lot of work by a lot of organisations who have had to in a sense break into this type of conversation.  And the key point I think for me is if we are able to articulate this in a set of -- in a language that we can at least agree on in about how we are describing governance and the issues at stake, then I think that is a necessary precondition for actually enables those further conversations.

     I think the first, one of the papers in the first session really showed how close some of these debates are.  I think that's, that issue of both transparency and shared language is something we have to address to broaden out these substantive concerns.

     >> Thanks for the question.  It's a fascinating discussion we have here.  What we can see, we had a discussion on that a couple of weeks ago in Paris at the jurisdiction conference.  There was this point, how to include people normally not included into the conversation.  One major issue.  And as regards transparency, the feeling there was that when it comes to some of the intermediaries, there is already transparency about at least the action of others.  The states, for example.  But there is some room for improvement when it comes to explaining redress policies, for example.  One Delegate there said if you do not know somebody in the institution, in the organisation or in the pressure group, you are lost.  You do not really have the opportunity there.  I think that is one basket of questions when it comes to transparency.

     Nevertheless, the main feeling was that it is mainly the states that lack transparency here.  There should be some good practices as regards transparency as regards state actions in this field.  And maybe we will come up with a kind of pilot project there with one or two countries, at least we have the intention to do so.

     So I think that is one aspect.  Another aspect I wanted to raise and that comes back to the initial point of Nick, the questions we are dealing with are extremely fundamental when it comes to legitimacy and rule of law and aspects like that.  Because as lawyers we normally have this simple binary code.  We say it's the state, the state actor, state action, it has to be a process of legitimization, laws apply or do not.  Then maybe the state is responsible for making others accountable, but they are not accountable as such.

     I think this very fundamental thing is this con settlement still valid?  Can we work with that when it comes to online platforms?  Or do we have to think about hybrid concepts where you look at specific actions and look at specific decisions and see whether they are at least de facto binding for people.  And then have a conversation about rule of law and things like that.  It is extremely challenging.  I have the honour to work with a Council of Europe group where we want to come up with suggestions on human rights.  Then that is for states and intermediaries.  When we come to intermediaries we have to ask the question:  What is the link here?  Can we really as Council of Europe based on Article ten of the European convention on human rights make some suggestions that we believe the actors are bound by Article 10?  What is actually the concept?  I think we have to put some thoughts and effort and research here.  I am looking forward to work together on that.  It seems to be very much linked, very many links between our research there.

     >> THIAGO TAVARES:  Any other questions?  Maybe we have someone in remote participation?

     Michael, I have a question for you.  One of the hot topics regarding corporate and policies, it is content blocking, including notes and take down of domain systems.

     In Brazil, for instance, we have a very strong discussion at the Congress regarding some Congressmen that want to make some change in the Civil Rights framework that we passed into law two years ago.  The copyright holders and especially the motion picture association of America, they are lobbying the Congress to modify to impose blocking on content and on domain name systems.

     I would like to know your views, if you could make some comments on that, especially on the impact of that kind of approach on net neutrality and freedom of speech.

     >> MICHAEL DICK:  Thank you very much for the question.  First, I think it is important to acknowledge that countries like Brazil, I think in my view and probably your view as well, have a much saner approach to Intellectual Property that I think better balances the public interest.  When you think about the regulation of patents and pharmaceuticals and drugs, Brazil I think is a very positive example of how to do this in such a way that it is actually both beneficial to those who are trying to make their living but also not doing so at the expense of broader social issues.

     I think that on the whole things like take-down notices and things that operate in that vein are to some degree less a violation of free expression because that is a very relative concept that tends to shift depending on what country you are in.  Some countries like Canada have clear limits to that.  Some like the U.S. don't.  It's up to courts to basically determine that right from the get go.  I think it is less about that and more about issues of liability.  The take-down notices, it is envisioned in DMCA and the new Canadian copyright regime and many other countries, they look at it as a way for ISPs in particular to avoid liability.  Sort of a safe harbor provision.  Look, we passed on the notice and our hands are clean.  It is a way to disintermediate the intermediary and pass on the responsibility for dealing with the content to the actual end user.

     The benefit to that, if there is a benefit, is that in theory, yes, you can sue the 12-year-old kid who posted the file.  Indeed the RIAA in the U.S. did just that.  The flip side to that, though, they don't typically have a lot of money.  So you can obviously try, but I think there was enough outrage with that sort of case, I'm not expecting to see as much of that personally.

     You still hear cases of that.  Friends of minor maybe yours who are getting random letters from MPAA or other random studios.  You have been downloading too much says the ISP of this person's content.  Go and take it down.  Otherwise you face legal consequences.

     In some respects that is totally ridiculous in the sense that there really is not a measurable harm economically, I would argue at least, being done.  So to me it is less about the issue of free expression because many countries recognize limits to that.  It is about trying to actually understand where the harm is actually being done.

     That is why I'm concerned about things like making available right and other things in that vein.  You basically come to a conclusion that is this problematic before it necessarily has even become problematic.  To win and infringement case you are supposed to prove there are actual damages.  Look at the damages and now let's get an order against this or find someone or what have you.  This is a case where arguably a legislative framework basically can say right from the get go:  No, someone exercised this right.  You can't do this.  They said this is infringement.  Or you have to pay a royalty just for doing this.  This very action of putting it out there, regardless of what actually happens to it.

     That I think concerns me more.  I mean, there is a time and place for issues of infringement or a take-down notice.  That is a separate conversation, a good but separate one.  It is a before the fact policing of the sharing of content that I think has a crucial role in the Internet Governance community because it is basically de facto regulating not just speech and expression but just what we can and cannot do with a server or medium or something like that.  I don't know if that quite answers your question, but I think it's an important perspective.

     >> THIAGO TAVARES:  Any other questions or comments?

     No?

     >> AUDIENCE:  For Wolfgang, if I might ask a question, it's another methodological question.  I'm interested in how you deal with questions of both scale and scope, massive decisions that are made here and scope about how you articulate or delineate the boundaries of these governance systems.  I guess the question is, can you understand the governance in the right to be forgotten without understanding conflicts over taxation policy?  And how do we then start to look at across these different issues of concern that involve some of the same but some new actors.  Can you reflect a little bit on how you are approaching that or how you think we might start doing that in the future.

     >> MARKUS OERMANN:  Yes.  This is Markus Oermann.  Regarding your question to the scope and how we can differentiate between the different discourses on governance or different governance issues, we try to apply the frame of relevance and we try to figure out how the actors in our figurations themselves describe the issues of governance.  How they themselves differentiate between different issues of fears of governance, different thematic frames in which they construct the norms that are relevant for the case is and regarding the large number of cases here, that's right, it is -- we are faced with numbers in 100,000s or millions.  But we try to get the abstract view on it by combining different methods, looking at it from the structural level and just asking for what are the cases when norms are, where there are frictions in the applications of norms.

     What are the cases that we can identify where the application is not so easy and these are the ones that are relevant for us because we are thinking that there are critical moments when governance structures develop further.  These critical moments are the ones that we have to identify.  We try to do this, but yes, it is a hard task to do it.

     >> WOLFGANG SCHULZ:  Maybe two more thoughts.  One just reflecting what Markus said.  Norms are there to deal with great numbers.  We are not interested in how the standard cases are dealt with but we are interested in when they there are changes of those rules and what can we see when we analyze this tipping point that there is a new rule or an amendment of a rule.  So that's one thing to deal with, scale and scope.  Of course, you are right, but I think we want to focus very narrowly on understanding specific backgrounds of how norms evolve.

     We do not see the whole picture or all of the motives of the actors, for example.  You are quite right, there are intersections between different sections of law, for example.  And of course, you can understand the action of some European politicians only against the background that they see that the money is flowing to the U.S.  And they talk about liability and other things, but that is what they have in mind.

     But I think that is a different research question.  An important one as well, and we want to engage in research around that in the next couple of years as well.  But this is more focused here.

     >> THIAGO TAVARES:  Any other comments, questions?  No?

     So I think we can move forward to the close.  Maybe if you want to speak a little bit again about your research, what is the next steps?  If you could briefly explain what is the next steps and how this community can contribute?

     >> NICOLAS SUZOR:  For me the next step is easy and incredibly difficult.  To understand all these questions web talking about and working with as many different people, trying to develop collaborations between all of us that have these shared research interests so we can start to understand and take on this massive task of understanding these questions at the different levels.

     >> PATRICIA VARGAS-LEON:  Well, I think the next step is, now we are seeing kind of reborn of the nation states after the Brexit and the last electoral results of the United States.  I can think of these regimes, a passionate call for these treaties, these institutions that are trying to turn back the power from the stakeholders to the governments.  I hope to focus and find possibilities of sharing research about other potential agreements that can also be covered by nation states, whether related to the air or the land itself.

     >> MARKUS OERMANN:  For us the natural next step is to do the project or to process it further and in my opinion, what is the next step in thinking about all these points that we touched on today is to do more analysis on the different power structures in the field.  So what is the difference between state power and exercise of state power, and the exercise of power by private actors?  And I think that is the point that all the papers have in common.  We could reformulate this question, what are the public interests here at steak and when does the private issue become a public issue?  That is the natural next step to these questions.

     >> WOLFGANG SCHULZ:  Coming from the four-factor model of governance research, not only do we need to study the doing aspect, what we do in this project or we try to see how the procedure works of norm making, but we also want to better understand the interaction between the social and technical norm-setting.  We have just started a cooperation with our Department of Software Engineering where we have a bigger three-year project on what we call information technology governance.  I always mix it up.  I think that's it.

     What we try to do there is to have a common understanding about terms and concepts, which is not so easy, of course, in a multidisciplinary environment.  We even want to come up with some technical solution if possible for privacy or other issues when we form these think tanks with different perspectives on Internet Governance.  Again it started a week ago.  So next year results, I guess.

     >> MICHAEL DICK:  It's Michael Dick for our online viewers.  I guess I get the last word.  That's scary!  Let me say a couple things.  One, on my end in terms of my work it is a question of continuing to write and research and hopefully turning more of this into active policy interventions.  I think that on that point, one thing that I would raise sort of in summation of a lot of the things we have heard today, there is a central theme here and central question of who gets to govern and how it is the process sort of legitimizes itself throughout the various actors and various things that are going on.  We've heard that from a multitude of perspectives.

     I think certainly that from my perspective looking at Intellectual Property policy, this is to me Internet Governance, just like a few years ago I was looking at technical standard setting as Internet Governance in the sense of why does the W3C basically call the shots?  This is the same thing in the sense that legal frameworks that obviously have their purpose are starting to encroach upon how we conceive of the online environment writ large.  That to me is alarming.  And that to me is something that I think we all need to be in tune with as we continue to do our work that ultimately will hopefully shape policy in its various forms, to keep in mind that central construct, that central question of who gets to write the rules and why is it that some actors in a political, in the broad political economy of the Internet and the web, et cetera, digital networks more generally, how is it that some voices are rising to the top whereas nonusers or those who are maybe not being heard as much are being left out.

     >> THIAGO TAVARES:  It's okay?  Yeah.  This one is working.

     Okay.  So we are closing this session.  We have 30 minutes extra for the lunch break.  And we, I would like to remember that we should come back at 12:30 for the next session.  It will be titled actors and policies in Internet Governance.  I would like also to recall that all the papers are available online at the GigaNet Web site.

     So thank you very much for all of you.  Congratulations again for your very interesting research.  Congratulations on your presentation and we will see you all after the lunch.  Have a good lunch.

     (Applause.)

     (The session concluded at 12:30 p.m.)