WEBVTT 00:00:00.000 --> 00:00:15.840 So in case you haven't heard about fanfiction yet, fanfiction is literature that is created 00:00:15.840 --> 00:00:22.440 by fans for fans without a commercial interest in mind. And so usually when we talk about 00:00:22.440 --> 00:00:29.880 fanfiction, two of these aspects come at the forefront, intertextuality and community. 00:00:29.880 --> 00:00:35.840 You can define these types of intertextuality with different levels of aggression, I would 00:00:35.840 --> 00:00:43.640 say, because Henry Jenkins famously called fanfiction writers textual poters. And then 00:00:43.640 --> 00:00:49.840 you can also say, well, they just use copyrighted material or they develop already existing 00:00:49.840 --> 00:00:57.800 root texts. And Abigail de Reco, for example, says that they are writing archontic literature, 00:00:57.800 --> 00:01:05.320 which describes this intertextual relationship more softly and without being so hierarchical, 00:01:05.320 --> 00:01:12.920 like if you use the words derivative or appropriated literature. And of course, then again, we immediately 00:01:12.920 --> 00:01:20.760 come to the community. So fan texts are created within a community and often for the community. 00:01:20.760 --> 00:01:27.060 But then in the next sentence, usually you will read about the quantity. There is a lot 00:01:27.180 --> 00:01:34.540 of fanfiction out there. And this is also the reason, according to some scholars, for 00:01:34.540 --> 00:01:42.340 example, Francesca Coppa, that why fanfiction hasn't been picked up by literary studies 00:01:42.340 --> 00:01:51.620 yet, because there's just too much of it and it's hard to focus on one thing in particular. 00:01:52.260 --> 00:01:59.540 She actually composed a fanfiction reader with sample fanfictions to provide some baseline 00:01:59.540 --> 00:02:06.940 for research. And she writes about it. It's there in the corner about this book. She writes, 00:02:06.940 --> 00:02:12.420 the most perverse thing about this book is that I'm giving you one Pringle, one Dorito 00:02:12.420 --> 00:02:19.380 and one Oreo, a single nacho, one pretzel nugget, a tiny little test a teaspoon of ice 00:02:19.380 --> 00:02:25.020 cream, one pita triangle worth of hummus. Then she of course says that you're not supposed 00:02:25.020 --> 00:02:32.060 to eat just one Dorito and goes on to say that the experience of fandom, especially 00:02:32.060 --> 00:02:37.860 in the age of the internet, is one of binge reading. Most new fans upon discovering it 00:02:37.860 --> 00:02:44.860 gobble it down. And even in the 1990s, so this is digital fandom, but even in the 1990s, 00:02:45.140 --> 00:02:54.140 Camille Bacon Smith wrote that first experiences of fandom are measured in inches or stacks 00:02:54.140 --> 00:03:02.520 rather than in titles. And this is when you had to acquire fandoms via zines that are 00:03:02.520 --> 00:03:09.520 somehow redistributed to you. So we have these three aspects, intertextuality, community 00:03:09.680 --> 00:03:16.680 and quantity. And why I find this important in the context of the social system of literature 00:03:19.360 --> 00:03:24.800 and digital methods is because we all know that there is this triad of author, reader 00:03:24.800 --> 00:03:31.800 and text. And of course we also know that there are other key forces like archival institutions 00:03:32.800 --> 00:03:39.800 or publishers or libraries who all deal with canonization. And Heidebrandt and Winkow describe 00:03:40.720 --> 00:03:47.600 this entire social subsystem with its actions and institution, the social system of literature. 00:03:47.600 --> 00:03:53.800 And some even say that literary history focuses on either the author or the reader or the 00:03:53.800 --> 00:04:00.800 text in different periods of its existence. And I think that now when we are talking about 00:04:00.800 --> 00:04:07.800 digital social reading so much, we can say that today literary history would do well 00:04:08.520 --> 00:04:15.160 while focusing on the social system of literature altogether. And what better way to do this 00:04:15.160 --> 00:04:22.160 than with digital methods? So we have different processes in place, the role of the author 00:04:23.280 --> 00:04:30.280 changes, we have different financing models and traditional publishing has to ease the 00:04:30.960 --> 00:04:36.840 process up because authors can now distribute their texts differently, maybe through social 00:04:36.840 --> 00:04:43.840 media. Then we see that there are different models of authorship. We are used to the single 00:04:44.040 --> 00:04:48.520 author, because this is something that has been established through copyright since the 00:04:48.520 --> 00:04:55.520 19th century, but now other models of authorship are resurfacing within digital spaces for 00:04:55.520 --> 00:05:02.240 example. At the same time for fan fiction authors, it is quite hard to distance themselves 00:05:02.240 --> 00:05:09.240 from these older authorship models and this is why many fan fictions start with the paratext 00:05:09.240 --> 00:05:16.240 of how people are not going to use the characters for any kind of financial gain as a safeguard 00:05:18.560 --> 00:05:25.080 against copyright claims. And so we see that fan fiction is actually neatly fits into this 00:05:25.080 --> 00:05:32.080 history of authorship and inter literary history as well. So in order to look at the 00:05:35.160 --> 00:05:41.600 social system of literature, what I'm coming to you with is a relational corpus from the 00:05:41.600 --> 00:05:48.600 website fanfiction.de. fanfiction.de is a German speaking fan fiction website which 00:05:49.600 --> 00:05:56.200 was, which exists since 2004 and it hosts both fan fictions and what is known as free 00:05:56.200 --> 00:06:03.200 works like original works. Today it covers over 1.4 thousand fandoms and over 400 thousand 00:06:04.320 --> 00:06:11.320 texts. It also has around 60, the corpus I'm working with has around 67 thousand texts 00:06:12.320 --> 00:06:18.320 by over 22 thousand authors. Initially the idea was developed at the University of Darmstadt 00:06:18.320 --> 00:06:23.320 where we thought it would be great to capture fan fiction production during COVID and we 00:06:23.320 --> 00:06:29.320 were thinking of course we will gather the data over a few months and then the pandemic 00:06:29.320 --> 00:06:35.320 will be over and so we can go on with our analysis. Well the data collection continued 00:06:36.320 --> 00:06:43.320 until 2024 and now the corpus is accessible on request and there is also a range of meta 00:06:47.840 --> 00:06:54.840 data available on the authors and on the texts. So what do we know from this meta data? We 00:06:56.160 --> 00:07:03.160 know that fan fiction.de attracts significantly more female users than male ones. In fact 00:07:04.160 --> 00:07:11.160 women make up 64 percent of the authors while men account for only about 6.7. Almost 27 00:07:12.640 --> 00:07:17.800 percent of users don't specify their gender at all and over the past four years the number 00:07:17.800 --> 00:07:24.800 of authors that identify as diverse has grown by a third which has to do with legislation. 00:07:25.240 --> 00:07:31.480 As you can see from the graphic most authors are between 18 and 40 years old. These findings 00:07:31.480 --> 00:07:37.960 aren't particularly surprising because we know that women have been dominating the fanzine 00:07:37.960 --> 00:07:44.960 subculture since the 1960s beginning with Star Trek and even in the 90s it was approximated 00:07:45.800 --> 00:07:52.800 that those fans who write fan fiction are probably around 90 percent of them are women. 00:07:53.800 --> 00:08:00.800 Whatever male fans were doing they were not writing texts and scholarly literature on 00:08:02.600 --> 00:08:09.000 fan fiction connects this gender distribution to historical ideas of identificatory reading 00:08:09.000 --> 00:08:16.000 or to the higher, the more complicated ways for women to enter the literary field. So 00:08:16.400 --> 00:08:23.400 what we also have within this corpus is relational data. So there are reviews to more than 51,000 00:08:28.160 --> 00:08:34.800 texts and the meta data on the stories and their protagonists. So this review data can 00:08:34.800 --> 00:08:41.800 be used in various ways. We can either look at how user A has commented on this or that 00:08:42.800 --> 00:08:49.800 chapter of user B but we also could build a network out of it. We can connect the users 00:08:51.400 --> 00:08:58.400 via their comments to stories. So the corpus includes texts, data on the authors, information 00:08:58.800 --> 00:09:05.800 about the texts such as genre and main characters and relational data in the form of reviews 00:09:06.200 --> 00:09:12.040 providing all the components of a digital social system of literature. So when we take 00:09:12.040 --> 00:09:18.320 a look at all the fandoms on Fan Fiction de l'Eye you can probably guess what is going 00:09:18.320 --> 00:09:24.600 to be the most popular fandom there. It is Harry Potter. It is followed by several anime 00:09:24.600 --> 00:09:31.600 fandoms such as Naruto and One Piece and My Hero Academia and across various platforms 00:09:32.600 --> 00:09:38.600 JK Rowling's books have captivated amateur authors and at first glance it seems that 00:09:38.600 --> 00:09:45.600 no other fandom can reach this. But let's keep in mind although the dominance is really 00:09:46.400 --> 00:09:53.400 unmet it is still only 18% of the overall content. In an initial paper together with 00:09:54.200 --> 00:09:59.200 a former Downstead team we focused on the Harry Potter fandom due to its dominance 00:09:59.200 --> 00:10:03.800 and made some interesting discoveries. For example there were slightly more male authors 00:10:03.800 --> 00:10:10.800 within the Harry Potter fandom and they received slightly more reviews to their texts. But 00:10:14.600 --> 00:10:21.600 if we introduce network analysis into this we can map and navigate our network of Fan 00:10:22.600 --> 00:10:29.600 Fictions and maybe filter out some of the texts a little bit better. So network analysis 00:10:30.600 --> 00:10:37.600 is a widely used approach based on the idea that real world networks somehow share similarities. 00:10:37.600 --> 00:10:44.600 Brandes and Erlebach formulated it as follows empirical networks representing diverse relations 00:10:44.600 --> 00:10:50.600 such as linkages among web pages, gene regulation in primitive organisms, sexual contacts among 00:10:51.200 --> 00:10:57.100 Swedes or the power grid of the western United States appear to have maybe surprisingly some 00:10:57.100 --> 00:11:02.900 statistical properties in common. So the idea is that the architecture of complex networks 00:11:02.900 --> 00:11:09.900 somehow is similar and we can use statistical tools to analyze and explore these different 00:11:10.200 --> 00:11:15.240 structures. So either we can be looking at Fan Fiction communities or collections of 00:11:15.240 --> 00:11:21.760 texts or characters and novels we can still using network analysis identify key actors 00:11:21.760 --> 00:11:28.760 and groups of entities that share certain similarities. So a network usually consists 00:11:30.640 --> 00:11:37.640 of nodes and edges and usually we can see that for example user A or A has written some 00:11:37.760 --> 00:11:44.200 letters to B and we may be interested in the content but we may be also interested in all 00:11:44.200 --> 00:11:51.200 the neighbors of A and B or the relationships among the neighbors and the neighbors of their 00:11:51.200 --> 00:11:58.200 neighbors and while the content of the letters stays the same there's a significant difference 00:11:58.200 --> 00:12:04.200 whether these letters are exchanged in a group where everybody talks to everybody or whether 00:12:05.040 --> 00:12:12.040 they are connecting two groups of people that otherwise do not communicate and we can also 00:12:12.760 --> 00:12:18.240 encode different information into these networks. We can say oh well C wrote 10 letters to D 00:12:18.240 --> 00:12:25.240 but D only replied once and then we can assess the different network positions who is more 00:12:25.240 --> 00:12:30.360 central who has more connections or who is connecting groups that otherwise would be 00:12:31.360 --> 00:12:38.360 not connected. And so you can also look at different clusters of actors and consider 00:12:41.440 --> 00:12:48.440 that something will be in common and in a previous project I examined interactions on 00:12:48.960 --> 00:12:55.960 an online forum of World War II enthusiasts and military historians and I could identify 00:12:56.160 --> 00:13:03.160 that when there were some sort of dense subgroups of communication on that forum then these 00:13:03.720 --> 00:13:09.440 different subgroups were representing different practices of memory. Some were interested 00:13:09.440 --> 00:13:16.440 in artifacts, other were interested in visiting bunkers and so on. And on the right you can 00:13:16.960 --> 00:13:23.960 see the texts also group according to the narrative according to first person narrative. 00:13:26.040 --> 00:13:33.040 So much for networks. This is how a network of our corpus looks like. So this is Fanfic 00:13:33.040 --> 00:13:40.040 2 and E and we can see that two users are connected by an edge if user A has written 00:13:41.160 --> 00:13:48.160 a review for a text by user B. And of course we also have metadata on the networks and 00:13:49.160 --> 00:13:56.160 this network is colored according to dense subgroups within the network. It consists 00:13:59.760 --> 00:14:06.760 of 47 communities ranging in size from 13 members to almost 4,000 and we have five largest 00:14:07.360 --> 00:14:14.360 clusters in the middle. These are approximately 40% of all users. So we can also map where 00:14:14.360 --> 00:14:21.360 do they interact within these subgroups using their metadata and then these are for examples 00:14:24.360 --> 00:14:31.360 all the subcommunities that deal with anime. And these are subcommunities that are somehow 00:14:32.440 --> 00:14:39.440 exchanging stories on Harry Potter. And here is Marvel and Star Trek and probably if you 00:14:40.440 --> 00:14:45.440 already did something on fan fiction you won't be surprised that Twilight and Fifty Shades 00:14:45.440 --> 00:14:50.440 of Grey are grouped together because one used to be a fan fiction of the other. And this 00:14:50.440 --> 00:14:57.440 is all quite not really surprising but this data driven approach enables us to see communities 00:14:59.440 --> 00:15:04.440 that we wouldn't have seen otherwise. So we all expect that there will be some exchange 00:15:05.440 --> 00:15:12.440 on Harry Potter but would we expect that the German series in Aller Freundschaft und Bergretter 00:15:13.440 --> 00:15:20.440 also would have a very tight knit and active community? I did not. Neither did I expect 00:15:20.440 --> 00:15:27.440 that German pseudo documentary criminal series from the early 2000s were also constituting 00:15:28.040 --> 00:15:35.040 a quite active community on fan fiction. And so I could go on here because there is also 00:15:35.800 --> 00:15:42.800 the medieval rock bands and fan fictions on Joko and Klaas and all sorts of things. But 00:15:44.640 --> 00:15:51.280 the next step here would be to look at how the interpretations of different texts within 00:15:51.360 --> 00:15:58.360 these sub communities differ. And what I could see and this is here we are coming to the 00:15:58.640 --> 00:16:04.560 changes that are happening within different stories. We have for example these two, the 00:16:04.560 --> 00:16:09.640 green cluster and the pink cluster which are mostly dealing with Harry Potter and what 00:16:09.640 --> 00:16:16.640 I noticed about them was that they have very very different distributions of gender. So 00:16:17.200 --> 00:16:23.240 we remember there are only 6% of male authors on the website in general but within the green 00:16:23.240 --> 00:16:30.240 cluster there are suddenly 16% of male authors. And the pink cluster only has less than 1% 00:16:32.200 --> 00:16:39.200 of male authors. So probably there is something more to it when we are looking at these differences. 00:16:39.200 --> 00:16:46.200 And so we were talking yesterday during coffee about how when you engage in a certain method 00:16:47.880 --> 00:16:54.880 becomes a hammer with which you can then go to the next step and create a next network 00:16:56.720 --> 00:17:03.720 because the data on the main characters within these two groups of people who exchange their 00:17:03.840 --> 00:17:09.800 views on the stories is also available. And so what do we see here? We see the protagonists 00:17:09.800 --> 00:17:15.520 of the pink cluster where mostly women are writing and of the green cluster that has 00:17:15.520 --> 00:17:22.520 quite a lot of quite a lot of male authors in comparison to the whole website. So we 00:17:24.480 --> 00:17:30.720 see that the green cluster features Harry Potter as the main protagonist and the pink 00:17:30.760 --> 00:17:37.760 cluster features Hermione Granger as the most central protagonist. However, their closest 00:17:38.200 --> 00:17:45.200 connections in both cases are to Severus Snape and Draco Malfoy. For in the pink cluster 00:17:46.280 --> 00:17:52.680 also quite interestingly is McGonagall who is connecting important characters from the 00:17:52.680 --> 00:17:58.800 canon but that do not have any connection for Hermione or any relevance Dumbledore or 00:17:58.800 --> 00:18:05.800 Voldemort are not very particularly interesting for Hermione but they are connected by McGonagall. 00:18:07.000 --> 00:18:14.000 And so we could either argue that these are two different ways of continuing to write 00:18:15.000 --> 00:18:21.360 Harry Potter. Either we take in a feminist perspective and place Hermione in the center 00:18:21.360 --> 00:18:28.160 of the narrative and some fan fictions do that, some do not. Then the other perspective 00:18:28.160 --> 00:18:35.160 is that Harry Potter is at the center and has more encounters with ambivalent characters. 00:18:35.600 --> 00:18:42.600 But it could also be that they are both used to tell more stories about Snape and Malfoy. 00:18:43.200 --> 00:18:50.200 And if we compare this data to data from other fan fiction portals then we can see that Draco 00:18:50.280 --> 00:18:56.560 and Harry and Draco and Hermione are very very common pairings. But Hermione and Snape 00:18:56.560 --> 00:19:03.560 and wait a second, Hermione and Snape and Harry and Snape are not that common. And I 00:19:04.120 --> 00:19:10.080 was wondering, this is data from Archive of Our Own which is the largest, I don't need 00:19:10.080 --> 00:19:16.120 to explain I think, is the largest fan fiction portal there is. And I was thinking maybe 00:19:16.120 --> 00:19:22.400 these are regional differences. Maybe this is something that German fans do and others 00:19:22.560 --> 00:19:29.360 do not. After all the Wattpet paper by Pianzola and Rebora show that there are regional differences 00:19:29.360 --> 00:19:36.360 in naming, conventions and so on. And then, and here I come to Middle Earth, I stumbled 00:19:36.600 --> 00:19:43.600 upon work on Vera Kunstlein and she famously demonstrated that in fan fiction fans tend 00:19:43.600 --> 00:19:50.600 to prefer younger characters in comparison to older ones. And this is not necessarily 00:19:52.640 --> 00:19:59.640 a regional question but I decided to check her data on Middle Earth to the data from 00:19:59.640 --> 00:20:06.640 fan fiction. Because of course we also have a Middle Earth cluster. And the Middle Earth 00:20:06.640 --> 00:20:12.600 cluster has around 1000 cells and it's a very large cluster. And I was thinking that 00:20:12.600 --> 00:20:19.600 there's around 1600 readers and authors. Middle Earth, here I combine The Hobbit and 00:20:20.640 --> 00:20:27.640 The Lord of the Rings, there are no fan fictions on Silmarillion for some reason. And it's 00:20:35.520 --> 00:20:41.800 approximately half of it. What I'm also interested in is how fandoms are combined because there 00:20:41.800 --> 00:20:47.400 is this idea by Henry Jenkins that we are all digital nomads and it's only natural to 00:20:47.400 --> 00:20:54.400 like different books and not just one. So here next to Middle Earth we have Merlin which 00:20:55.440 --> 00:21:02.440 kind of is self-explanatory, it's also fantasy, it's also set somewhere in the past probably. 00:21:02.480 --> 00:21:08.280 But also interestingly Kalmais Vinnitu is also quite present there. And we can look 00:21:08.280 --> 00:21:14.680 at it in a little bit more detail. Here you see a model of this cluster in the form of 00:21:14.680 --> 00:21:21.680 a bi-model network. So we can browse and explore the community using network characteristics. 00:21:22.720 --> 00:21:29.720 We can take a look at what is the most central text and then just navigate our reading and 00:21:31.080 --> 00:21:38.080 we don't have the problem that there is too much fan fiction to even start. The triangles 00:21:38.160 --> 00:21:43.960 are colored, the darker ones are Lord of the Rings and the brighter ones are The Hobbit 00:21:43.960 --> 00:21:49.040 and then Kalmais and Merlin and they are located at the center because it's so tightly mid 00:21:49.040 --> 00:21:56.040 and then Kalmais and Merlin are up there. Yeah and so what you see in Middle Earth fan 00:21:57.720 --> 00:22:04.720 fiction as well is that Hobbit lends itself more to classical fan fiction production. 00:22:05.720 --> 00:22:12.720 So this is a very very common way, very common distribution of genres. It's a love story 00:22:13.560 --> 00:22:18.840 and then a little bit of drama but also family and romance and romance again because there 00:22:18.840 --> 00:22:24.480 are two labels of romance in German and in English. You still haven't understood how 00:22:24.480 --> 00:22:30.320 they are differentiated and there is also the classical fan fiction genre, the hurt 00:22:30.320 --> 00:22:37.320 and foot genre which means that this is a type of story where one protagonist gets 00:22:42.040 --> 00:22:49.040 injured and is nurtured back into health and is a very common trope in fan fiction. But 00:22:50.120 --> 00:22:56.960 in Lord of the Rings you have a very different situation. You have adventure and drama and 00:22:56.960 --> 00:23:03.960 fantasy and no romance at all. So it sounds to me as if the Hobbit writers would stick 00:23:07.320 --> 00:23:12.560 more to fan-ish kinds of productions whereas Lord of the Rings fans are more stuck to the 00:23:12.560 --> 00:23:19.560 original. And I did not annotate the occurrences by the books and this is based on the German 00:23:20.560 --> 00:23:27.560 original because the fan fictions are of course also in German. And we can see that users 00:23:27.560 --> 00:23:34.560 on Fan Fiction DE much prefer to use the story world of Tolkien rather than the original 00:23:34.560 --> 00:23:39.560 characters. So on the left you can see the top ten protagonists which is basically the 00:23:39.560 --> 00:23:46.560 fellowship and the middle table demonstrates the same metric but shows the occurrences 00:23:46.560 --> 00:23:52.040 in the fan fiction corpus. While the right table demonstrates the number of main characters 00:23:52.040 --> 00:23:59.040 within the corpus. So people occur in fan fiction but then also they use tags to navigate 00:24:01.040 --> 00:24:08.040 the reader experience. And what we see is that Hobbits are on decline and more and more 00:24:09.040 --> 00:24:16.040 of elves and humans are also used. And the other thing that we see is that Hobbits are 00:24:16.560 --> 00:24:23.560 more popular. When we compare this to AO3 data we see a very different preferences. 00:24:23.560 --> 00:24:29.360 Aragorn is more popular than Legolas and Gandalf disappears from the list completely which 00:24:29.360 --> 00:24:36.360 supports the argument that older characters are excluded in favor of more conventionally 00:24:38.720 --> 00:24:44.720 attractive ones. Although older in the context of Lord of the Rings is a whole different 00:24:44.720 --> 00:24:51.720 story. As is the case for a lot of fan fiction the main tendency in the genre on AO3 is erotic 00:24:56.840 --> 00:25:03.360 in nature. So in the case of Fan Fiction DE it seems that the character tags sort of work 00:25:03.360 --> 00:25:09.280 as mechanism of expectation management and perhaps stories with the fan favorites like 00:25:09.360 --> 00:25:16.360 Legolas and Thranduil attract more attention from users. And the strong difference between 00:25:16.440 --> 00:25:21.680 these two platforms is an argument against this theory of a common fan fiction narrative 00:25:21.680 --> 00:25:28.680 that fans sort of retell the story into one direction. And whether this is language, culture 00:25:28.760 --> 00:25:35.480 or platform specific remains very much open ended. And finally when we look at the text 00:25:35.480 --> 00:25:41.000 of Lord of the Rings compared to the original in a very simple contrastive analysis I'll 00:25:41.000 --> 00:25:45.560 say a few words about that you don't have to read it all. We can see that there are 00:25:45.560 --> 00:25:51.520 some differences between the two that demonstrate how despite writing in the same genre still 00:25:51.520 --> 00:25:58.520 means something quite different. In the list on the right we have quite a few landscape 00:25:58.520 --> 00:26:05.080 descriptions which Tolkien is quite known for. One or two character names Frodo who 00:26:05.080 --> 00:26:12.080 is not mentioned at all in the fan fictions for some reason occurs within the original. 00:26:13.040 --> 00:26:20.040 And then we see and this is why I love digital analysis we see many many many references 00:26:20.200 --> 00:26:27.200 to light and darkness. And this is something that the list on the left which is the list 00:26:27.840 --> 00:26:34.240 of preferred words of the fan fiction doesn't pick up on at all. So on the list on the left 00:26:34.320 --> 00:26:40.840 we have plenty of interactional verbs more informal style references to the body lips 00:26:40.840 --> 00:26:47.640 hair shoulders and so on. And then lots of narrative time references which I'm also not 00:26:47.640 --> 00:26:54.640 sure what to do with. Somehow this may be due to the fact that amateur writers need 00:26:56.720 --> 00:27:03.720 more telling than showing which comes when or maybe this is because the narrative in 00:27:04.440 --> 00:27:11.440 Lord of the Rings is a little bit more straightforward than here. But overall it raises the question 00:27:11.800 --> 00:27:17.980 whether fan fiction has a style of its own. And in conclusion we see that we have different 00:27:17.980 --> 00:27:24.760 interpretive communities and they have different dynamics across platforms. They can be subversive 00:27:24.760 --> 00:27:30.040 but they don't have to be. And then we have the digital paratext that serve as a guiding 00:27:30.040 --> 00:27:37.040 mechanism for readers and shape their experiences. And finally and my most important takeaway 00:27:39.400 --> 00:27:45.480 message is that with network models you can definitely provide some structure to the reading 00:27:45.480 --> 00:27:52.480 experience and navigate the corpus much much better. So that's all.