 I'm Heather Endershot, director of graduate studies for comparative media studies. Today, our speaker for our colloquium series is Jens Polman, who is a research associate at the Center for Media Communication and Information Research at the University of Bremen. He received his PhD from Stanford University in 2017 and his research centers on internet policy discourse in both Germany and the United States. And his first book, the creation of an avant-garde brand, Heinrich Mueller's self-presentation in the German public sphere, will be published this fall, congratulations. That's great. The title of his talk today is platform regulation and the digital public sphere, comparing the discourse in Germany and the United States. Take it away Jens. All right, thank you very much for the kind introduction. I'm very excited about having the opportunity to present my work here today. So to let you know, I am a visiting scholar here at MIT actually since October, I think. So I've been here for six months or so and I enjoyed it very much. Kurt Fendt is hosting me and I had the pleasure of taking some classes with him. So I met with undergrads from MIT, also together with him and two Europe's were working on this research project that I will present here today. So I'm very thankful for that as well. And I can say that I've been working with the libraries very much. So AJ Tornader has been helping me and us, so the team to deal with access to data resources and all this has been tremendously helpful and exciting for me. So a big thanks to MIT and to all of you. And the only thing I didn't work out so well, actually because of the pandemic was meeting people and reaching out to you guys. So when I came, it was still kind of difficult and now I see it's getting better but now I'm leaving already next week. So that's the only thing that didn't work out, I guess, as planned. But today may be a chance to present some of my work and get into discussions with you and potentially we can continue some of that in the future. So let me see, let me share my screen so that, yeah. So I guess now you should be able to see my slides. So that's the title for today. I will jump ahead and give you the Gliederung, that's Germans, that should be the structure, right? That's the roadmap. First thing is a short introduction about this topic of the talk today but also generally my research. Then I will talk a bit more in-depth about the Network Enforcement Act. So a short version is NET-STG, so a particular anti-hate speech law in Germany that I think is very interesting. Also for an international perspective and a transatlantic perspective on platform regulation. Then in point three, I will give you some insights on the research that I have been doing in the last kind of two, three years on this NET-STG working with a text corpora analyzing this discourse. I wanna show you what we have done and zoom particularly in for this over blocking analysis. So what is over blocking and why have people talking about this with regards to NET-STG so intensively that is, if you will, that's the main part of, or the main argument, main questions that I wanna put out here in this conversation. And then the last point is research that I have done or that we are doing here at MIT. So I wanna talk about the US side of things. So that's section 230 of the Communications Decency Act. I guess that will be brief, but I wanna say what we have done here and also talk about access to data resources for text data mining, because that's what we have been involved with, particularly in the last couple of weeks. So that's the roadmap, let's jump right into, so platform regulation in Germany and the United States is the main topic, why is it important? I think the big question is, so who can and actually who should determine the structure of the evolving digital public sphere that is the big question and it is particularly relevant when it comes to this net-stg law, I will explain it later, but as I see it or I guess you can at least take it this way that net-stg and the content moderation practices of platforms with their community standards or as community guidelines as they sometimes say, these are competing regulatory systems that shape what we can say on these platforms. And by doing that, they also shape what can be said in the digital public sphere and what the digital public sphere actually consists of. And obviously that is very important and it is so important because the public sphere and the digital public sphere lays the foundations of forming political opinions, both in everyday interactions as well as on election day. And it's so crucial obviously for democracy and Western liberal democracies in particular. And what you can see here is this cartoon that I sometimes like to show. So you have the net-stg on the left side and big platforms on the right and the middle is it's us potentially, right? The users and we are, you know, poked at I guess at least here. And it seems that sometimes we get into the middle we're sandwiched in this kind of fight for power over speech. And I think that in my research, one of the key issues is trying to find ways in which we can empower ourselves, I guess, you know, the users and in which they can have as much influence as possible and can make the shape of this digital public sphere as whatever democratic as possible. So that's what I think is important about my own research. So let's talk about this network enforcement act. So net-stg, I have a couple of slides with texts. I'm sorry about that, but whatever, you know, give you, so I'll talk and then you can additionally read if you want to. The point is this net-stg was enacted in July of 2017 and then it came into effect in January 2018 so a little transitional phase. And there was a very controversial discussion about this law both in the Bundestag, which is the German parliament, as well as in the media in Germany, but also in the United States, at least among scholars and internationally. To give you a bit of a historical context, I think it's important to be aware of what happened before this law was enacted. Obviously, we have the Brexit referendum in 2016, as well as the U.S. presidential election in 2016. Cambridge Analytica became known and people were aware of the fact that at least potentially democratic decision-making could be influenced by the using, let's say, of online platforms and basically manipulating potentially what voters may want to do or want to say, so at least that was there. People were thinking about this at the time, so it was seen as a danger that this political decision-making process could be hijacked by using social media platforms. There are two other particularly German things that were important. On the one hand, there was a big election coming up in 2017 and people were afraid that similar things as happened in 2016 could happen in Germany, as well. And there was this shift to the right in Germany after the European refugee crisis. You may remember Angela Merkel opened the borders. People were welcome from Syria, but this also led to the right-wing party, at least so there's one in Germany called the AfD, AFD that got more votes and became stronger and people in Germany were very concerned about that. So this is just to give you this general context, I would say. There's another thing to be aware of and that is European lawmakers had been talking to the platforms for a couple of years before that. They have said you need to act quicker on terrorism and on a hate speech and take this content off and platforms generally have said, oh yeah, sure, we'll do that and we'll look into that and then not much really happened. So that is I guess the historical background I would say. And what happened then, so the German lawmakers decided to enact this law. What is it actually all about? It says that platforms need to have a system in place that enables users to flag content that they deem to be unlawful. And unlawful in this case means that it is unlawful based on the provisions of the German criminal code. Strafgesetzbuch, that's the German word and that has paragraphs as defamation, incitement to hatred, insults and so on. So it's about I think 22 or 25 paragraphs that you could flag content for with regards to the German criminal code. Platforms would then have to remove reported content. So content being reported by users that is in quotes manifestly unlawful and do that within 24 hours of receiving notice of it. That is of course putting pressure onto these platforms 24 hours depending on how you see that can go by very quickly. Generally it has been considered to be putting lots of pressure on them. But as you see below in more difficult cases they were also allowed to take something like seven days to discuss these issues and make a decision or they could also outsource it to an independent regulatory body. So there were ways of getting away from this time pressure. But generally speaking time pressure I would say was there and it's a main element of this law. In addition the platforms would have to publish reports about the incoming complaint. So how many pieces of content have been flagged according to net CG and then what did they do about it? Did they decide to take it down? Did they do that in the given time frame or not? Or did they leave it up on the platform? They would have to publish reports about that every six months. And the last piece is that platforms that are mostly all of them I think are from the US. They also would have to establish a legal representative in Germany that the government or other institutions could reach out to and basically talk to. Because that was not given. So these platforms were having tremendous impact on the public sphere in Germany but there was no one that actually the government or anybody else could reach out and say hey there's something happening or whatever because there was no office or whatever that was actually representing that. And that changed by the net CG law in 2018 for sure. The big problem with this law are the fines. So on the one hand they being forced to react to content very quickly in 24 hours but then also the fines. And here's the stuff on the fines if platforms failed to implement such a functioning system of flagging contents or if they systematically failed to take down content that is unlawful or was deemed to be unlawful in the given time frames if they would not publish the required reports then they could get fined and that with penalties up to several million euros I think 50 million euros. So that was a steep fine but something that is important here and I highlighted it in the slide is systematically oftentimes in the reporting and also that happened with scholars all over the place but also from the US oftentimes it seemed as if they are thinking if there's five pieces of content or whatever then the fine would be leveraged on these platforms that is not the case systematically is important so that a systematic failure to comply with these three points would actually bring about those fines. What is the relevance of that? What's the important thing about this net CG law? I think what is important is it is the first time that a Western liberal democracy so as Germany enacted this kind of a law with regards to unlawful speech that actually undermines US intermediary intermediarily liability law sorry that was just too much but basically that was given in section 230 CDA was undermined by this law in the sense that platforms are partially held liable for user generated speech of which they're usually protected from by section 230 so as you saw there they're only held if you will accountable if they don't do these things that I just mentioned so if they don't have the system up if they don't react systematically to things and if they don't report stuff but it is a first step at undermining section 230 that is why it's so important especially since it happened in a Western liberal democracy and the other point and that is more my point I think net CG is so important because of the first point but then also because if you look at it and look at it more closely you see that there are different approaches with regards to platform regulation but also generally values with regards to free speech obviously democracy and then the digital public sphere to look at them, to be aware of them and to discuss those issues also with regards to then coming to a joint stance I would say on that of how to proceed in the future with the US and Europe here that's a post from the Human Rights Watch so in their opinion this law was very problematic so here they're saying that's the wrong way to deal with online abuse and it should be reversed so criticism was harsh particularly I would say or among other places from the US what was this criticism here I just have a few points so there's more to that but I don't want to bore you too much but important is surely the first point so net CG may lead to the privatization of your radical decisions particularly here then regarding free speech so it's the platforms who decide what you can say or what you cannot say and not a court which is democratically legitimized so that's first big point the second point is over blocking we're going to talk more about it but the idea is that this law would encourage these platforms actually to err on the side of rather taking stuff down instead of looking closely to it not because they're afraid of the fines so that is one major piece of criticism generally of course you know danger to the free flow of information and the internet that was mentioned a couple of times and then also the last point copycat laws by authoritarian regimes it actually happened that Russia drafted a law that was very similar to the net CG and of course they are using it in very different ways than Germany does but the point is surely valid that nation states that don't have a democratic tradition like the United States or Western European countries can use this boilerplate to do things that are not very democratic so I guess that's fair to say there were a few points or arguments in favor of net CG several I would say one is a very German take on things actually since Second World War in the Holocaust the German basic law the Grundgesetz is actually based on the idea that sometimes you have to restrict basic democratic values such as freedom of expression, privacy also the market in order actually to protect democracy against extremists that's a take away from the Third Reich and in Europe in January France I think this exists as well so this belief is fundamental at least for the German constitution and so that's what they're saying you know it's fine to have such a law because it actually helps them to protect democracy that's the idea behind the other thing mentioned often in Germany is that if these platforms operate in Germany they should actually comply with German law I guess that's also something that people can understand very well obviously there are cultural differences with regards to content I mean the easiest one is Nazi symbolism is not allowed in Germany in any way it's been taken down very quickly and you get sued if you post any of that stuff which is not the case in the United States in some other countries so that is a cultural decision I would say that is very important to us but it's different in other countries other thing maybe doesn't always have to be these horrible examples but also let's take nudity Facebook to down a lot of pictures of breastfeeding women in Germany or Scandinavia people just shook their heads nudity is not a problem there or that sort so they don't understand actually but the issue is there cultural differences they come into play with regards to content regulation and it's at least something to think about should there be should platforms react to these cultural differences or not or must they even and then the last point is that received very positively was that having this legal entity that represents these platforms in Germany or in these nations that was actually applauded I think across the political board we thought that would be very important and I think I mean I think it's an important point and it's also obvious that the market power that Europe has and also Germany is a reason why platforms were willing to do that and there are many other countries and other parts of the world where this is not happening just because their market doesn't generate so much leverage okay so that was my ideal if you will on the general question of this law and what's at stake now I want to take you a bit closer at my research and tell you what I have done to analyze this course about it and what we did is the first thing is we put together a digital text corpora first thing is the first corpus is based on German IT blocks so those are bloggers and websites that report about tech related issues often times tech and policy also sometimes products we scraped that from the web and made it accessible with the Brunnenberg Academy of Sciences and Humanities and if you want to look at that link so there you can access this data I want to just show you here a bit so the kind of blocks we have there are particular IT law blocks we have intellectuals, professors whatever writing about these kinds of questions also blocks that talk about tech products of different sorts and then what I consider to be very interesting is IT news outlets so some blocks developed actually from these kind of immature individual blocks into more professional blocks so they have real staff they report actually every day about developments in tech and policy and they are actually the main thing that got me interested in blogs or IT blocks in the first place because they and other bloggers may actually have a real impact on the discussion that is out there in the public sphere and also in the more traditional media because they are experts and they are monitored for whatever they post so this is the first corpus we searched for our search terms all these terms here in German relate obviously to this next DG law and this was then so that means we took all blog posts that mentioned one of these search terms regarding next DG at least once and put this into our corpus, our sub corpus what you can see here is the percentage that individual blogs then the share that they have in the sub corpus so how many articles that actually they have that mention this particular law and you see nettspolitik.org that is one of those IT news outlets that develop from amateur to more professional and they are very very involved in this discussion about the law and other ones as well but then also they are smaller ones so this is basically the setup of bloggers talking about this particular law what you can see here is how many article or how many blog posts did these individual blogs contribute per month to the publications about this particular law don't want to get too deep into that but what you can see is by this the publications basically follow the political process so as you may remember in June 2017 the law was enacted this around this is where you have the first peak and then it came fully into effect in January 2018 so this is when most people blog about this stuff is not so surprising but it is also interesting that you see a discussion then drops but it continues basically for a whole 2019 so it still is relevant continuously second corpus that we put together and that was a little more difficult was daily German newspapers we were able to compile text for the same search terms with regards to net-stg from those nine most important German newspapers they really cover I would say everything that is of importance with regards to nationwide German newspapers so very excited to have them all together in one place and what you see here is again the setup so how many articles did those individual newspapers contribute to sub corpus and the second thing here is again how many articles did these individual papers contribute per month same thing here is what you see they follow the political decision-making process so June 2017 and January 2018 being the two peaks but the discussion still kind of continues on afterwards here you have the two graphs together so blog posts and newspaper articles basically follow the same pattern here what we did then is so I was so interested in this question about over blocking we took these two corpora and we said okay now let's look how this discussion evolves how this course evolves with regards to this over blocking thesis there were others I mentioned the privatization of theoretical decisions or other things you could also have looked at those ones but I was particularly interested in that one so this is the one that I'm picking right here over blocking again is this idea so it actually is interesting so Germans think that the word is English right or American it's not I think here it's over removal or over regulation or just censorship but you know Germans think it sounds so nicely English and I think it's not a bad creation maybe it's going to find its way back into English language at some point but anyway over blocking the idea is that it's a science that the next law imposes or threatens the platforms with in the end leads them to delete more content than necessary to avoid those fights right simple simple idea also makes a lot of sense and that would then of course restrict people's freedom of expression and would be a bad thing interesting is that there's another thesis out there is what I call the anti thesis and that is basically the opposite of it and that says because it is in the the platform's economic interest to keep users there as long as possible so that they can gather that data so that they can sell advertisement to them and for these reasons they would actually not you know just remove stuff you know without you know looking closely at it just arbitrarily they would rather actually try to have a nuanced deletion policy in order not to you know get their users annoyed and they're potentially leaving the platform or not contributing content anymore and you know that obviously being this favor to their own business model so this is the second thesis and I would say it is interesting at least to have it in mind and I'm going to say this already here since there is no we don't have enough data to actually know what the platforms are doing because the platforms are not sharing their data there's no access to objective data about about their about their taking down of content I think we do not empirically know if the anti over-blocking thesis or the over-blocking thesis is true we can only have theories about it but there is no empirical way to decide that and that's why I was thinking it's interesting to look at them both and see how intensively they are represented in a discourse or not and so that's what we did I won't get too technical but what we did we took these to corpora we classified for some examples with the Max QDA analysis so that's like Atlas IT mixed method working environment then we trained machine learning model with these examples we ran this model over the corpora so the IT blocks and the newspapers joined together this throughout hits passages where the computer thought this is the over-blocking thesis or the anti over-blocking thesis and I went through that manually and decided is that truly one of these thesis then I categorized it that way or if it's not then it was taking out of the equation that's what we did and what you can see here is a representation of that in blue you see the over-blocking thesis and the amount of articles that mention this thesis over time so everything that is blue is the over-blocking thesis and you can see that's all over this chart it's actually almost completely blue and the orange is the mentioning of the anti over-blocking thesis and that only appears very few times I actually checked the last one there in 2019 is not really relevant so you could even kick that out but it's somewhat obvious that the over-blocking thesis is much more prevailing here this chart does the same thing but it's just relative to the amount of articles published articles and blog posts per month so it's a relative number but it stays basically the same so over-blocking thesis shows up many many times and throughout the whole time frame that we're looking at here so from the first discussions about up until it's been being in place for I don't know what two years or something like that you can I think summarize that very easily the over-blocking thesis is presented very very often 290 or almost 300 times while the anti over-blocking thesis is only showing up 7 times so that is a tremendous disbalance and that is I would say interesting so it's obvious this over-blocking thesis dominates the discourse and I think it's interesting because of what I mentioned before we in my opinion maybe it's up for debate but we cannot empirically determine which of the two is true and just because we don't have access to the platforms data and that is why actually I would be thinking as a researcher or as an academic student that one would at least mention this anti over-blocking thesis once in a while to say well it could actually also be that the platforms will do this in a nuanced way because it's in their own interest so when and when we started this I was actually expecting the anti over-blocking thesis to be much more present at least a bit at least whatever 20, 30 times or so so only 7 that was surprising to me as well the other point that I mentioned down at the bottom is it's also interesting because the sitting German minister of justice and several researchers in Germany as well supported that anti-theses and particularly the minister of justice obviously has the platform to make his views heard so it's interesting that that wasn't taken up neither by the IT blocks nor by the newspapers right and that's just I mean I just think it's interesting how is that why is that happening so this leads me to those questions that I also see in the realm of a critical discourse analysis so the question is what's the reason for this balance where is it coming from and who are the stakeholders or potentially interest groups that help back so this one thesis and why have they been so successful actually in communicating this thesis over the other because you can say I mean yeah and almost 90% of all articles about this topic there was this one sentence or two that says oh you know there's the danger of over blocking happening due to net-stg so I just think that's interesting and also with the last point so I mentioned about justice it also seems not to have pursued the strategy of focusing on this anti-thesis very much and that's also I think a fair question to ask what did they do seems they went for a different strategy so they pushed rather different sorts of information but is also I think interesting again also from the academic point if they are really if it's really at least difficult to find which of the two is true then why is there hardly any mentioning of the anti-thesis yeah so that is basically my the questions that I'm working with and I'm working on based on those results that just showed I'm happy to share some insights on where I think this is going or what happens here but in the future with regards to my research but I also wanted to give you an update what happened in the last couple of years because right so the NET-STG has been in place since 2018 so that's four years and a bit it's important I guess to say that no finds were ever issued according to the NET-STG so that just simply never happened that's also important because I think it shows that the law was rather intended to put pressure onto these platforms but not to find them and gain any money out of that I guess of course you couldn't know that in the beginning but looking back at it now I think it's fair to assume so the other thing is sure also I mean yeah the end of the free internet in Germany did really not come about I would say it's still possible in Germany to freely express your points of view also because they have just not been many reports about over-blocking or stuff being repressed in Germany at all so there are a few instances right in January of 2018 when takedown decisions were whatever didn't work out so well I would say there were little scandals at the beginning but I'm really talking here three or four five instances but apart from that never anything else came up so I think it's very fair to assume that no major scandals have been there because the blogs and also the media that I looked at they would have been really happy to jump on that and make it a big thing and it simply haven't right and then you can see that when you read these blog articles you can see that netspolitik.org was sometimes trying to make a story out of this one artist being blocked or whatever but you just look at it and say well that's not really that's I mean it's always unfortunate if any sort of speech gets taken down or repressed for the wrong reasons but those examples don't make a case right so that was I think interesting and is important to mention there also has been an update to the netstg maybe I don't go into this right now I could talk about it later they didn't change too much they made adjustments improvements I would say but let's not go into this but rather let's look at the takeaways from this whole procedure or for this whole experience I would say what you can take away I think are these things on the first hand netstg has started a very much needed and necessary debate about the power of platforms and about the impact that they have on the structure of the digital public sphere I think that is that's definitely the case also in countries other than the United States obviously in Europe that's a different thing than in the United States so I think it really has helped to think about these issues and to put them onto the plate you could also say there are several I think really positive aspects I didn't list them here but what Facebook did and other platforms is they hired people in Germany to go through content based also with German language skills being aware of the cultural situation in that state or in that nation also with access to a better education for their job and psychological resources for dealing with their job as being content moderators because surely you know the discussion about outsourcing this kind of work to third world countries and having those people deal with this psychological stress that comes with looking through all kinds of content that is on platforms so they are at least in these content moderation places they have some resources psychological support and so on and so forth I think that is definitely an improvement and I think also that actually is an outcome of the next TG and also it's good that these platforms are more present now in Germany that a discussion also with policymakers is there is more likely it seems they are reacting more to what people care about so those are good outcomes what I consider to be really interesting from the theoretical perspective are these next points are technically part the point too so there is a US tradition in talking about over blocking or here Daphne Keller would rather say over removal is talking of new speech regulation and the idea that third party speech platforms are rather willing to just take down speech that is not their own but belongs to third parties in difference to let's say newspapers so this is a strong tradition and very important original research if you take these results that I just presented today the question really is do they still apply and to what extent do they apply or do they have to be readjusted or not or is it also a European question is it saying that next TG may not have lead to over blocking if one says that and agrees on that let's just do this for an experiment thought experiment if you agree on that then the question is is the theory that assumed it would lead to over blocking or to over removal is it still valid or not that is I think one of the most interesting questions although again that's what I said in the last point so is the next TG experience an exceptional case also an exceptional European or potentially also German case that is also questions so could you say this theory still applies in the United States or other countries but here in this particular case in Germany in Europe it just was exceptional and there are things that are somewhat different this could also be something or what could also be the case is that actually that kind of theory tradition is still right we just have noticed yet that actually there was over removal due to I don't think that's very realistic but it's you cannot I would say it's too early to make an absolutely positive claim on that so this is also where I'm working on from a more theoretical perspective I think that is why an SDG so interesting because yeah it this question will it lead to to over removal is just so pressing and so interesting and engaging with this kind of thinking alright so this already leads me slowly to to wrap up with this last point that is the research that I have done or I am engaged with here at MIT and that is me trying to compare the discourse about an SDG with the discussion about a reform of Section 230 as I already mentioned the big question to me is so has the discourse actually changed since the events of the storming of the capital in 2021 so last year so Section 230 you know has been discussed I would say as the same with the net SDG mostly among experts right it's not people on the street are not usually meeting in the pub and talking net SDG or Section 230 that's sure but my case is that net SDG became a thing after it was enacted and it started a politically interested discussion also among you know broader swaths of the population I mean not so broad but you know a little broader and what I was wondering is has something similar potentially happened after January 6th of last year also with regards to Section 230 also the question is is it now different kind of stakeholders who talk about Section 230 or not are they bringing up different sort of arguments when they talk about it in the past there was there was always you know stuff related to free speech in Section 230 obviously but then there was also copyright concerns and then there was sex trafficking as a big topic I was wondering will those events of the past year have shifted discourse maybe a little closer to what this discussed with regards to net SDG so democracy as a whole extremist behavior behavior and those kinds of topics that would be of interest to me also to see then can you really compare those two discussions well or can you not so well because they're obviously two different kind of laws Section 230 is much broader and much more basic and much more in a sense also important and net SDG is much more specific and deals with also just a German situation so the question is can you compare it well I think you can and I'll be working on that but it's still I guess up for debate so what we're doing here or what I've been doing in the last couple months is trying to set up digital text corpora in order to do actually more than the same stuff that I told you about with regards to the German context now just in the US about this other kind of law so we're putting together data from Twitter and from Reddit from IT blogs that are based in the United States or international as well as we're looking at academic journals and US newspapers in order then again to build sub corpora that are just about this discourse on Section 230 with some you know search terms and then see okay who are the people that get involved with these discussions so who are the stakeholders what are the arguments that they bring forward what are the topics and so on and so forth stuff that you can do with text data mining and here I think that's one of the last slides that I have important maybe for people that are interested in this kind of a digital discourse analysis with text data mining is the point that it was in the past years always very very difficult to get a grip on current or contemporary newspapers to do text data mining on that right so you could get access through FACTIVA or Nexus Lexus these providers though oftentimes you know send out PDF files or whatever so computing this stuff was difficult because then you would have to change the format into something that's better readable by machines and so on and due to copyright restrictions and so on and so forth but now something has happened something has changed and that's at least for me the big news so with Consulate and ProQuest actually data providers have decided to bring to to produce environments that allow you to work on their full text data so with Consulate it's JStore articles and articles from Portico with ProQuest it's newspaper articles and journals and other data that you can work on in an online environment basically doing cloud computing where you can work on the full text so you can do computing on the full text but you don't download it so you can do that without infringing copyrights and without these newspapers in your times that you take the data and then post it somewhere online and then people won't read the newspaper anymore so this is a very important tremendous step because I can just tell you doing the analysis in Germany with these nine newspapers what we had to do is at the German National Library they set up a machine for me a computer that was hooked off the internet so it didn't have access to the internet there they put these files for me and I could work on them but again separated from the internet and only on the premises of the national library so it was because of these copyright and license agreement restriction it was very difficult and very tedious to work with these kinds of materials and what Consulate and ProQuest are providing to us right now is for a researcher like me absolutely fascinating and exciting and that's also why I'm so much in touch with AJ from the MIT Libraries because I think these opportunities are just truly fascinating and so what we're doing in this little research team I mentioned so I have two Europe's that are working with me and Kurt on this we're doing these trials with Consulate and ProQuest with ProQuest we have been working already for some time so they have as I mentioned they have the data available online they have Jupiter notebooks which allows you to run your Python or R code and do your analysis and then you can download metadata or you know graphs that belong to that so we're doing this right now and actually today we have gotten access to ProQuest which does the same thing just with all these newspapers and it's I mean I'm just I'm stoked as you would say in California stoked about those you have everything there you have Washington Post, Washington Times New York Times Post and US Today and the Wall Street Journal and so many different resources from different political spectra I guess that was never possible before so that is that gets me very excited and I think is important oh yeah so I already talked about this so you do the calculations on a cloud computing environment and then you can do topic modeling named entity recognition all kinds of other analysis and that is very convenient and very easy it's also good for working with students so these these two students doing this with they already know coding but it's also to teach first steps so I think these environments are great and I also do believe and maybe that's my last point that this kind of a setup will boost text data mining research on these contemporary resources like these media newspaper articles so I would if somebody from MIT asked me if they should continue this trial with ProQuest and Consulate I would say totally you should absolutely do that I think that researchers that are interested in this kind of stuff will always want access of the sort and I think that universities that have it and can provide it to their researchers will have a head start and will be able to produce interesting research and so I would say if you can then you should do it and I will see if my home institution if the University of Brim will be able to do something like this as well okay I think that was it from my side I went a little over time but hopefully not too much so thank you very much for listening to my stuff I'd be very excited if you have any questions for me also reach out if you want to with this email address or contact me on Twitter Wow, thank you so much Jens, I'm completely stoked about the work as they say in California about the work that you're doing I've decided questions over here but I want to see if I don't see anything in the chat right now do we have any questions out there from maybe to start with our graduate students Yes, Tomas Thank you and it's great to meet you and this is extremely exciting and I wish we would have known each other before because I'm interested in the same topics and I think it's wonderful that you're looking at platforms that are super interesting work that I don't see nothing I was really interested in the fact that you were looking at in this case in Germany where people are the idea is that it's a mechanism to implement the law but what I've seen that discourse around the US is that the law is not enough concerns around, for example, vaccine misinformation there are not things that are, per se, illegal is that as if the law was not enough in the case of the US whereas I would see in Germany you're looking at a completely different concept so I was wondering if you have thoughts about this came up in your research I feel there's like two different imaginaries of the law that are involved Can you be a little more specific, where do you see the difference between the United States and Germany in this case? It's very vague but I would say that in the US the concern, for example, by the government is that, for example, vaccine misinformation is being spread in social media platforms or that hate speech is spread on social media platforms things that are not illegal they are legally permitted by the First Amendment but the concern would be that these things are antisocial but not illegal whereas I would say that the phenomena you're tackling in the case of Germany is things that are they know how do we apply the law in this realm Yeah, I guess you're pointing to a very interesting issue that is obviously the traditions are so different because the First Amendment is such a knockout argument there's nothing that you can do about it so the traditions are so different that it doesn't make sense, I assume, also to go at it potentially from the legal perspective because it would be tremendously different to set up something that is really similar and that is already so I had this question up there and you've given an answer to that question, you're saying basically it's so different that in Europe it may be working in different ways so there may not be overblocking in Europe and the United States it may be different but the thing is that one thing that I at least see and where I wonder is what happened at least with President Donald Trump is that in the end he was the platform so his speech rights were tremendously restricted and also it happened by these for-profit firms basically and by no means by any sort of democratic decision involved and I'm always wondering if that is if that is a better solution to this kind of question or not or again the answer from a European perspective is obvious Europeans would probably say no that is not the good way to do it there should be some democratic involvement and there should be some government impact on that but also here I'm not so sure I mean how the US population or whatever thinks about this or how the legal setup is my assumption is that this may be different in the United States it may be there the idea is rather that for-profit companies doing that is actually what they feel potentially more comfortable with right but I think this is exactly the point to think about and to continue conversations and to think through it but also in a sense that I wonder what it means if these systems are so different that you know potentially as I outlined it here theory that is based on the US media landscape and platforms may just not apply to Europe in this way if that is the case then I wonder if European nations and researchers should still take this for granted also the theory and that is something that I just don't- let's say I'm thinking about I don't know if that's a good answer to what you've been asking but No thank you so much I wonder if one huge difference and I don't know maybe it's not but between the European context or German context in the US context is the kind of symbolic way that children are used in the discourse of censorship and censorship around speech and the notion of protecting children so that for example in communications law here the notion of indecent speech which is a violation of FCC policy is not allowed to protect children so you can have indecent speech after during a safe harbor for indecentity that is late at night when it's assumed that children are asleep that sexually or excretorily specific speech is legal right but it's not legal you know at 6pm when children are soon to be part of the audience so it's like children become a catch-all to enable a certain kind of regulation in the US and I'm wondering if they are used in a similar symbolic way elsewhere and I guess I would also well this is a separate issue but I mean there are certain kinds of speech that are legally censorable in the US that like defamation has certain laws around it that speech that is censorable in certain contexts or obscenity is technically you know illegal speech that inherently does harm like yelling fire in a crowded movie theater right so just wondering so it's two part question one is about the children childhood issue and protecting childhood innocence and the other is about the kinds of speech that have long been protected in the censorable in the US and whether there's a comparable body of censorable speech in Germany outside of some of the things that you've outlined well I would say the discourse about an SDG did not touch very much about protection of children in this regard as you mentioned and as it's prevailing in the United States it was very much a discussion about these kinds of you know speech rights with regards to protecting democracy there was also so the and that is I mean the government actually one position that they pushed for is saying that they do this to protect minorities and minority speech because they are usually the ones whose voices are pushed to the margins if the mainstream if you will you know speaks up and uses hate speech or a speech that is at least you know offensive or whatever then usually those on the margins are even pushed further out they don't speak up they leave for whatever and this is why we want to push this harder that these communities have more of opportunity speak up and actually in Germany it is the LGBTQ community and the Jewish community has continuously pushed for even harsher laws and also this update about the SDG so they have been really trying to actually make this law also lead platforms to report unlawful speech to the government in order that then it would take been taken to court right away so I guess it is rather been discussion about these kinds of minorities and how you can protect them from not being pushed out of public discourse I would say definitely then so children's issues have been to my understanding even they have been hardly been in this discussion interesting I mean it is such a huge difference I think from the U.S. context and also just the kind of status of say what it means to be an extremist in the U.S. and from a legal perspective versus in Germany you know we said specifically their argument in favor of SDG is that you restrict democracy protect democracy against extremists in the U.S. context I thought what would be the argument in the national debate about what an extremist is and it would be quite different than I would think then in the German context. Interesting I mean that would be interesting to me as well I mean or to look at who is actually considered an extremist in both countries would be interesting I mean to your second point maybe and I hope I got it right I would just say interesting to me is regards to the law that so you have these paragraphs right according to the German criminal code but I wonder if I mean if I were to interpret the move to actually you know enact this kind of law I would really think and guess that it was not so much about individual cases of these paragraphs but rather of pushing for the attentiveness of these platforms to cater to whatever in this case German needs and be more responsive and be more aware and these kinds of things I mean I so I think at least in the first kind of couple of reports these six months reports was often mentioned in defamation and some sort of insults they turned out that the platforms didn't take too much down in anyway right so they looked at it but then oftentimes they made a decision this is covered by free speech and we just leave it up there but yeah again I got the feeling that pushing them for more action was really intended again with this fear that you know happened in this particular situation of 2017 and being forced to you know react Shrushi I see you have your hand up Hello thank you so much for your lovely talk my question is just to get your general take on the conversation of decentralization and how that might be a place of research or the sort of move of even capital into nations that are considered decentralized like there's no one governing body is there a potential to challenge some of the perceptions of platforms and between different continents just general thoughts on this is a trend it's a buzzword right now it's used everywhere in finance and governance so what do you think of it well you know I don't know if I have a good grab on that but I would think that trying to make money off of the platforms as the social media platforms do you know brings a centralization with it you know that is basically the opposite of decentralizing them and people just throwing in stuff and that actually their content moderation practices and these guidelines are actually the opposite they are structures that try to keep you know bring things back to the center or at least allow them to make those decisions and to structure it if you will then discourse or I mean it's important for them truly in a sense that they want to create these communities that they want so they can stay there but I guess what I'm saying is indifference to a truly open platform that would allow people to contribute in ways that they would want to and that would be truly decentralized this kind of closed if you will platform this structure that these kind of platforms build up is opposing I guess this ideal I don't know if that is what you've been asking for but if you throw that buzzword at me then that would be my take on it I'm sorry I muted Heather oops sorry about that do we have any questions from our attendees who are outside of our main room I'm just looking over here to see all right well I will other I have a few more that I've written down here surprise you ask specifically you know has the discourse changed since January 6th and I think that you started to answer that but I wonder if you could answer that more and it gets back to my question about extremists and so on and or by extension issue about protecting children and so on and so far so many extremists from January 6th believe that there is a sex trafficking ring load by Democrats and you know that the Democrats are the extremists and they're the not extremists and so on but anyway could you unpack a little bit how the discourse has changed since January 6th from the U.S. and at least or is January 6th understood in a German perspective even like you could respond from both sides well that's a couple of interesting things there so I mean I would say from from our analysis we are just not there to make claims about what has changed since January 6th because we just don't have the data yet to say anything about it so I couldn't make a clear statement on that my assumption you know would be that the discourse has has grown in the first case I mean that's not so surprising potentially but that is what I assume and as I already said I assume it drifting away from these other issues to a more politically focused and if you will then extremist a question but I mean it really would have to to look at it more closely I'm also wondering will more you know more mainstream whatever outlets or users whatever have engaged in that conversation I assume so but also you know there's no data there that we could I mean I could base my claims on so I think will I mean I would rather wait then to make any claims that I cannot back up in any way from the German perspective I mean I I didn't do particularly research on it but I mean Germans have been so critical of the last administration in so many ways from the right the beginning and it is just I mean you get I mean I was in the US by the election if you're German you cannot not think of 1933 I mean it happened to everyone right you were immediately you were afraid and anxious and of course that maybe not you cannot compare history that easily but that has been so prevailing in Germany from the outset and in this in this regard also I think that Germans have been very relieved you know when there was a takeover of a range of power that was more or less peacefully and they were I mean I guess you know shocked about what happened there and how things unfold I mean that's very general but this is how it has been it has been seen with regards to the platforms I think people were also interested but also shocked I mean I think the power of these platforms has become obvious to really now everyone what they can do and that they can shut down also a political leader and they also have noticed many times that you know things are so different now since Trump is not tweeting every day anymore and he's not on the news you know right and it's basically somebody pulled the plug and things have been completely different and I think that has been very very much noticed in Germany as well which again doesn't necessarily make the thing completely makes it easier because it means you have to deal with these entities and it seems that maybe people are happy that the European Union is pushing in many directions you know with privacy law so with this next approach you have now the Digital Service Act and the Digital Markets Act that tries to deal with these platforms but on the other hand what really matters is the United States in some sort of way and there I think people see or are assuming that there is hardly a way to really act on these platforms potentially well that's what they've seen you know people had to report to Congress or what I have been questioned by Congress but there's no substantial movement that could solve also easily solve that question so these firms are very very powerful they have a huge tremendous influence and one wonders how the U.S. will be able to whatever deal with that also from a political perspective one does wonder thank you Tomas has some questions for us I believe I was wondering as a MRSA not only the specific research product but more in your general research if you have seen different accounts of the digital public sphere in the U.S. and Germany and whatever that means right Yes I would say in so many ways I mean the media landscape is very different we have this public so öffentlich-rechtliche Medien which is public media broadcasting they are financed by the state they inject something of importance in a sense of a certain viewpoint or a variety of viewpoints that I, although I'm not so much of a media user here as I am in Germany but I am as well which doesn't seem to be present in this way so that is right there a tremendous difference and you know that also means right it is the market is less there is a little restriction on the market side because the government funds media in this sort of way which already changes the whole setup a little more so I mean these are very basic dimensions but also when you look at usership I think a lot more people in the United States take their news from social media than in Germany there is always rising in Germany so year by year but still there is a tremendous difference it also showed that in this particular election that I mentioned in 2017 actually there were hardly any reports about bought influences or whatever this kind of Cambridge and Odeika stuff that happened in 2016 I don't know if that's the case because people were more aware of that and where we're technically already fighting better or also because the media system in Germany is simply different people relying more still more on traditional media newspapers this kind of public broadcasting and media but I think you know it's fair to say that the impact wasn't as as intense as in the United States and Great Britain in 2016 so I mean those are just a couple of things that come to mind thinking of the public sphere but I'm pretty sure there are many others that come to mind but may I ask you maybe you're not born in the United States how is it in your country do you recognize a clear difference I mean yes I'm from Argentina I mean the public sphere is obviously different very radically different but I'm thinking that my question and I think you answered it partially was understanding specifically of how they see the digital public sphere right how do they see the realm of conversation on the digital sphere and on digital media and it's something interesting that happened very recently it's that the government that they were trying to start some conversations around what good practices for social media companies so they stop intoxicating our democracy right and I'm thinking of Biden's comment about when he was asked what did he think about Facebook probably related to vaccine misinformation he said that social media companies need to stop destroying our democracy or something very similar I think so I've been thinking a lot about how they understand different countries what it means to be a healthier democratic debate in a liberal public sphere on the internet I may just say two things what brought me also to this topic is that was the great idea so the internet would bring more freedom to all of us and more democratic discourse and so on and so forth and then we are all aware that maybe that's not really going to happen so much also because they are centralized again these platforms and they have an agenda and they're also as you mentioned potentially spreading fake news or polarization or polarizing content is useful to them and will bring more money in all of these things but I also thought one of the issues why I'm looking at this is that it's actually technical experts who have access to this and not so much former politicians, the traditional politicians and it's so difficult for I think the everyday Joe on the street to understand what's going to what's happening here from a legal perspective but then also from a from a technical perspective for a lot of for some time people didn't realize that their feeds are constructed in some sort of way and that is based on some sort of data and that you can be manipulated by the pieces of information that is set in front of you now people know especially of course when you are in media studies but for years ago people weren't necessarily aware of that so this research project is also about seeing how can you translate these technically and legally complex issues for a discussion in a society so that people can make informed decisions about their future and the way politics should be done and next DG discourse and potentially other ones as well are just examples where I wonder if that actually really works or works well and what I presented to you with the over blocking and anti-over blocking thesis is just one example I would say simply if you were to deal with this course you would mention these two options it could be this way, it could be that way and then people would look at research as what they said about it and then you would make an informed decision this is not happening for different reasons I haven't gone into explaining that too far but that's interesting for me to look at and I have at least a few ideas something that I at least still wanted to mention because I'm always such a European you cannot drop that but I just wanted to mention that one surely has to keep in mind that Europeans or the European Union also has an interest in diminishing the power of these platforms because they've never been able to come up with something on their own there's no European platform that provides, except for there's a Scandinavian thing, Klarna or so they do finances and then there's another one here but basically there are no big powerful platforms that provide people with those essential social media features that whatever, Facebook, Twitter and so on do and so there isn't a I think it's it's obvious that there's some political, economic incentive to push for more privacy legislation, to push for something like NET-STG also to say okay we want to have a say again we want to get a grip on the public sphere but also we want to reign in potentially these platforms that is I think both it is driven by truly important and democratic values but there's also an economic competition going on there I think it would be naive not to see this point I just wanted to mention that so that is also part of the picture that that has a component in the European approach to this thank you thank you thank you so much Jens I think that we're going to end here but I'm going to make a few plugs and announcements before we go, not just for our grad students but for other, for our guests one is that a week from today we're going to have a Finnish speaker, Oscar Wienberg who recently defended his dissertation on the 70s TV show All in the Family and he's really speaking from a kind of political historians angle in a way more than a TV studies angle and looking at the uses of this TV show to promote legislation for fundraising, for political candidates promoting agendas and so on and it's really interesting and quite unique work the day after that a week from tomorrow we're just going to pop this up into the chat area, a link here we have thesis day when our grad students present their work formally, their master's thesis to the community and MIT people can come in person, it's in Bartos theater and other people can come in person too if they get Tim tickets which we can tell you about you can email me or Andrew Whitaker so it is open to the public there's a reception afterwards and it's also streamed if people want to do that from a distance so I want to encourage everyone to attend that, it's really like our showcase for how wonderful our graduate students are and finally I'll just note I think in many ways relevant to what we've been talking about today on April 14 Martha Minow of the Harvard Law School will be speaking with me in conversation about her recent book Saving the News which is thinking through regulatory issues and how it's very dramatic title as she's thinking specifically about the death of print media and the ways that news media and the ways that regulations have sort of not kept up with technological changes and I'm certain that I haven't finished reading the book yet but I'm certain that section 230 is going to be a really important issue that we're going to be talking about so in many ways quite different methods from legal scholar from what we've been talking about today but I think it's going to be quite relevant to some of the issues we've been talking about today so with those plugs for the future I will end here and I will thank you again again and give you a round of applause for presenting your work it's really great to meet you while you are visiting here with us and thank you so much for sharing what you've been doing so perfect thanks so much it was a pleasure thanks