1 00:00:00,000 --> 00:00:04,360 [MUSIC PLAYING] 2 00:00:04,360 --> 00:00:07,400 Welcome, Silke Schwand, to Leipzig. 3 00:00:07,400 --> 00:00:10,600 Once again, not sure when you have been here the last time. 4 00:00:10,600 --> 00:00:11,100 I think-- 5 00:00:11,100 --> 00:00:13,000 For the historic attack last year? 6 00:00:13,000 --> 00:00:13,500 Yeah. 7 00:00:13,500 --> 00:00:14,840 Oh, historic attack, yeah. 8 00:00:14,840 --> 00:00:17,800 And also for a digital humanities event, 9 00:00:17,800 --> 00:00:19,160 I think two or three years ago. 10 00:00:19,160 --> 00:00:20,880 Yeah, that's true. 11 00:00:20,880 --> 00:00:22,920 Great to have you back. 12 00:00:22,920 --> 00:00:25,880 I know you very well, but I think 13 00:00:25,880 --> 00:00:27,960 this is not the case for everybody here in the room. 14 00:00:27,960 --> 00:00:32,000 So I will do what, well, is typically 15 00:00:32,000 --> 00:00:35,280 expected when important people like you are introduced. 16 00:00:35,280 --> 00:00:39,540 I will tell people about the most important stages 17 00:00:39,540 --> 00:00:42,400 in your academic career. 18 00:00:42,400 --> 00:00:45,080 So I think it all started in Bielefeld. 19 00:00:45,080 --> 00:00:47,880 So Silke Schwand is a trained historian. 20 00:00:47,880 --> 00:00:52,840 And I think the next stage of your academic travels 21 00:00:52,840 --> 00:00:56,760 was London, and from London to Frankfurt. 22 00:00:56,760 --> 00:01:00,240 I think it was also Frankfurt, where you did your PhD. 23 00:01:00,240 --> 00:01:01,800 And I think in Frankfurt, you already 24 00:01:01,800 --> 00:01:06,080 started to get more digital, because you 25 00:01:06,080 --> 00:01:09,980 started as a traditional historian back in the day. 26 00:01:09,980 --> 00:01:15,080 And we will see you digitized yourself and your research-- 27 00:01:15,080 --> 00:01:17,520 not yourself, but your research-- 28 00:01:17,520 --> 00:01:18,900 through the years. 29 00:01:18,900 --> 00:01:21,360 After finishing the PhD in Frankfurt, 30 00:01:21,360 --> 00:01:24,040 I think there was another research project in Frankfurt. 31 00:01:24,040 --> 00:01:27,200 And then you came back to Bielefeld. 32 00:01:27,200 --> 00:01:30,520 I think first as a [SPEAKING GERMAN] 33 00:01:30,520 --> 00:01:33,840 whatever that means in English, as a lecturer. 34 00:01:33,840 --> 00:01:36,800 And then you were appointed a professorship 35 00:01:36,800 --> 00:01:41,120 for digital humanities and medieval history. 36 00:01:41,120 --> 00:01:43,140 It's digital and medieval history. 37 00:01:43,140 --> 00:01:44,560 Now it's only digital history. 38 00:01:44,560 --> 00:01:45,480 Yes, that was my point. 39 00:01:45,480 --> 00:01:46,440 There is a difference. 40 00:01:46,440 --> 00:01:51,800 Showing the incremental digital digitization 41 00:01:51,800 --> 00:01:53,320 of the researcher Silke Schwand. 42 00:01:53,320 --> 00:01:57,560 Because I saw in your CV, like two years ago, 43 00:01:57,560 --> 00:02:01,480 you switched the denomination to just digital humanities. 44 00:02:01,480 --> 00:02:03,200 So no more-- 45 00:02:03,200 --> 00:02:03,680 History. 46 00:02:03,680 --> 00:02:04,760 It's digital history. 47 00:02:04,760 --> 00:02:05,520 Digital history. 48 00:02:05,520 --> 00:02:06,240 Sorry, I know. 49 00:02:06,240 --> 00:02:07,840 Sorry, I'm very particular about this. 50 00:02:07,840 --> 00:02:11,440 Still a historian, but a digital historian. 51 00:02:11,440 --> 00:02:13,960 And no more medieval history. 52 00:02:13,960 --> 00:02:16,080 Well, as you will see later, there 53 00:02:16,080 --> 00:02:17,600 is some medieval stuff going on. 54 00:02:17,600 --> 00:02:20,200 So I bet it is digital history. 55 00:02:20,200 --> 00:02:24,720 So that's for the very brief, well, academic career. 56 00:02:24,720 --> 00:02:27,640 As I said, I also know Silke from, well, 57 00:02:27,640 --> 00:02:30,560 the digital humanities community very well. 58 00:02:30,560 --> 00:02:33,160 Actually, there's a couple of projects I could mention. 59 00:02:33,160 --> 00:02:36,680 One of my-- I will mention my favorite project, which is-- 60 00:02:36,680 --> 00:02:38,160 I think it's called eTARDIS. 61 00:02:38,160 --> 00:02:41,800 I never understood what the acronym stands for. 62 00:02:41,800 --> 00:02:42,800 I will tell you later. 63 00:02:42,800 --> 00:02:44,120 It's in the presentation. 64 00:02:44,120 --> 00:02:44,960 OK. 65 00:02:44,960 --> 00:02:46,760 So can I spoil it right away? 66 00:02:46,760 --> 00:02:50,800 So it's about virtual reality for visualization purposes 67 00:02:50,800 --> 00:02:53,160 in history, which is an amazing topic. 68 00:02:53,160 --> 00:02:58,000 And yeah, you can expect some very nice demonstrations, 69 00:02:58,000 --> 00:02:59,040 I guess. 70 00:02:59,040 --> 00:03:04,000 But Silke Schwand is not only, well, 71 00:03:04,000 --> 00:03:05,880 an applied digital historian. 72 00:03:05,880 --> 00:03:07,880 She's also very much interested in the theory 73 00:03:07,880 --> 00:03:11,000 of digital humanities and digital history. 74 00:03:11,000 --> 00:03:14,400 And this is where we, I think, had our last encounter, 75 00:03:14,400 --> 00:03:17,200 collaboration. Well, collaboration sounds weird 76 00:03:17,200 --> 00:03:22,720 because it was a staged dispute between a historian 77 00:03:22,720 --> 00:03:25,080 and a computer science person. 78 00:03:25,080 --> 00:03:29,600 And we didn't really dispute about algorithms. 79 00:03:29,600 --> 00:03:32,880 And I introduced algorithms from a computer science perspective. 80 00:03:32,880 --> 00:03:40,920 And Silke argued that traces of algorithmity 81 00:03:40,920 --> 00:03:46,640 can also be found in humanities and especially in history 82 00:03:46,640 --> 00:03:51,320 research because you have procedural, very systematic 83 00:03:51,320 --> 00:03:51,920 workings. 84 00:03:51,920 --> 00:03:54,720 And maybe you will also come back to this in your lecture, 85 00:03:54,720 --> 00:03:55,880 I guess. 86 00:03:55,880 --> 00:03:59,720 Yeah, so many more important stuff to tell you about Silke. 87 00:03:59,720 --> 00:04:03,000 But I think she will present some interesting research 88 00:04:03,000 --> 00:04:03,520 herself. 89 00:04:03,520 --> 00:04:05,640 Maybe one last important thing, which 90 00:04:05,640 --> 00:04:08,280 is very recent in her activities. 91 00:04:08,280 --> 00:04:11,560 So Silke can be really happy that she's here 92 00:04:11,560 --> 00:04:14,280 because she's really busy right now because she's planning 93 00:04:14,280 --> 00:04:16,880 the digital humanities conference, 94 00:04:16,880 --> 00:04:19,520 the German digital humanities conference, DHD. 95 00:04:19,520 --> 00:04:21,400 Maybe some of you know it. 96 00:04:21,400 --> 00:04:26,560 And it will be in Bielefeld next March, March 3rd to 7th, 97 00:04:26,560 --> 00:04:27,200 I guess. 98 00:04:27,200 --> 00:04:27,880 Yeah. 99 00:04:27,880 --> 00:04:30,160 And yeah, whoever of you was involved 100 00:04:30,160 --> 00:04:32,240 in organizing a conference or even just a workshop 101 00:04:32,240 --> 00:04:33,800 knows it's a lot of pain. 102 00:04:33,800 --> 00:04:37,400 She's like the head of the local organizing committee. 103 00:04:37,400 --> 00:04:40,840 And it's really a pleasure to have you here, 104 00:04:40,840 --> 00:04:45,840 despite these efforts of the organization. 105 00:04:45,840 --> 00:04:49,640 And if you're ready for your talk, I will hand over to you. 106 00:04:49,640 --> 00:04:51,600 Thank you very much. 107 00:04:51,600 --> 00:04:54,600 Thank you very much, Manuel, for the kind introduction. 108 00:04:54,600 --> 00:04:57,600 And I kind of second the war invitation 109 00:04:57,600 --> 00:04:59,080 to Bielefeld next year. 110 00:04:59,080 --> 00:05:02,920 And the motto of the conference is under construction, one, 111 00:05:02,920 --> 00:05:06,880 because the campus in Bielefeld is very much under construction. 112 00:05:06,880 --> 00:05:11,280 And two, the whole discussion about digital humanities 113 00:05:11,280 --> 00:05:15,120 or digital history, whatever it is in this regard, 114 00:05:15,120 --> 00:05:16,880 it is our understanding it should be more 115 00:05:16,880 --> 00:05:19,440 about being under construction. 116 00:05:19,440 --> 00:05:24,320 So trying to build together approaches 117 00:05:24,320 --> 00:05:27,520 on different disciplines, a very formal approach from the computer 118 00:05:27,520 --> 00:05:31,520 science or computer linguistics, so a very formalized approach 119 00:05:31,520 --> 00:05:33,640 when you work with computers, with something 120 00:05:33,640 --> 00:05:38,880 that is very interpretive, very hermeneutic, very also proud 121 00:05:38,880 --> 00:05:42,120 to be vague and not trying to kind of pinpoint 122 00:05:42,120 --> 00:05:44,560 each and every single data point. 123 00:05:44,560 --> 00:05:47,440 So that's where the motto under construction comes from. 124 00:05:47,440 --> 00:05:49,840 And this could also have been the title of my talk today. 125 00:05:49,840 --> 00:05:53,920 But I decided to do something similar content-wise, 126 00:05:53,920 --> 00:05:58,120 but to start at a different point. 127 00:05:58,120 --> 00:06:02,040 So today, I will talk about figuring out the past. 128 00:06:02,040 --> 00:06:05,760 Do numbers tell stories? 129 00:06:05,760 --> 00:06:09,560 What you can see on the slide is the only hint 130 00:06:09,560 --> 00:06:13,280 at artificial intelligence that I'm going to do today. 131 00:06:13,280 --> 00:06:19,200 So this is a AI-generated image that you get when you ask-- 132 00:06:19,200 --> 00:06:21,400 in this case, it's OpenAI's Deli-- 133 00:06:21,400 --> 00:06:23,580 to give you a picture about figuring out the past. 134 00:06:23,580 --> 00:06:25,840 And I thought this was really interesting because it had 135 00:06:25,840 --> 00:06:30,400 all by itself the idea that it's about books and figures. 136 00:06:30,400 --> 00:06:32,560 But I'm not going to go and delve into discussions 137 00:06:32,560 --> 00:06:33,800 about artificial intelligence. 138 00:06:33,800 --> 00:06:36,360 We just had at the beginning, Manuel and I, 139 00:06:36,360 --> 00:06:39,160 a little conversation about how much we despise 140 00:06:39,160 --> 00:06:42,200 the term artificial intelligence in and of itself. 141 00:06:42,200 --> 00:06:46,760 I'm not going to open that box, not today. 142 00:06:46,760 --> 00:06:49,960 I will start today with telling you 143 00:06:49,960 --> 00:06:53,500 something about where the first part of the title comes from. 144 00:06:53,500 --> 00:06:56,000 It is a quote. 145 00:06:56,000 --> 00:07:00,080 It's the title of the book by Peter Turchin and Daniel Hoyer. 146 00:07:00,080 --> 00:07:02,100 It was published in 2020. 147 00:07:02,100 --> 00:07:04,460 And it is called "Figuring Out the Past-- 148 00:07:04,460 --> 00:07:11,160 The 3,495 Vital Statistics That Explain World History." 149 00:07:11,160 --> 00:07:13,080 And I have this book, and it sits on my desk, 150 00:07:13,080 --> 00:07:14,960 and I always wanted to talk about this book. 151 00:07:14,960 --> 00:07:17,220 So this is basically where my idea-- 152 00:07:17,220 --> 00:07:20,600 to make this lecture about the question, 153 00:07:20,600 --> 00:07:24,480 what is quantifiable in history, or what kind of quantifiable 154 00:07:24,480 --> 00:07:29,360 data can we use when we want to tell histories 155 00:07:29,360 --> 00:07:32,200 or when we want to build historical narratives? 156 00:07:32,200 --> 00:07:36,960 That's what my lecture is going to be about today. 157 00:07:36,960 --> 00:07:42,520 I quote from the introductory remarks 158 00:07:42,520 --> 00:07:45,360 by Peter Turchin, who himself is quite the accomplished 159 00:07:45,360 --> 00:07:47,280 historian. 160 00:07:47,280 --> 00:07:50,520 And he has a big project about collecting historical data 161 00:07:50,520 --> 00:07:53,000 that I'm also going to introduce in a short while. 162 00:07:53,000 --> 00:07:55,480 So back to the quote. 163 00:07:55,480 --> 00:07:57,480 "We are used to thinking about history 164 00:07:57,480 --> 00:08:01,840 in terms of stories, who did what to whom. 165 00:08:01,840 --> 00:08:04,960 Yet we understand our own world through data, 166 00:08:04,960 --> 00:08:07,080 vast arrays of statistics that reveal 167 00:08:07,080 --> 00:08:09,680 the workings of our societies. 168 00:08:09,680 --> 00:08:11,280 Why not the past as well? 169 00:08:11,280 --> 00:08:14,220 Figuring out the past turns a quantitative eye--" 170 00:08:14,220 --> 00:08:16,640 and figuring out the past is the book-- 171 00:08:16,640 --> 00:08:20,720 "turns a quantitative eye on our collective trajectory. 172 00:08:20,720 --> 00:08:23,920 Behind the fleeting dramas of individual factions and rulers, 173 00:08:23,920 --> 00:08:26,840 it looks for large-scale regularities. 174 00:08:26,840 --> 00:08:30,560 It asks how key social and technological innovations 175 00:08:30,560 --> 00:08:32,160 spread around the world. 176 00:08:32,160 --> 00:08:36,640 And it pinpoints outliers from general trends." 177 00:08:36,640 --> 00:08:40,000 And I think this quote really sort of opens up 178 00:08:40,000 --> 00:08:45,400 the space between two poles, the very quantitative, very 179 00:08:45,400 --> 00:08:50,200 statistics, very data-oriented interpretations that we maybe 180 00:08:50,200 --> 00:08:53,640 associate with social sciences rather than with history. 181 00:08:53,640 --> 00:08:58,760 And we tell stories about wars and people and interactions 182 00:08:58,760 --> 00:09:04,680 in a very sort of unclean kind of way that we, 183 00:09:04,680 --> 00:09:07,080 at least according to this quote, 184 00:09:07,080 --> 00:09:12,720 have a way to tell stories about history. 185 00:09:12,720 --> 00:09:15,080 You can see this in the behind fleeting dramas 186 00:09:15,080 --> 00:09:16,560 of individual factions and rulers. 187 00:09:16,560 --> 00:09:19,440 I mean, I think this is an image of history 188 00:09:19,440 --> 00:09:22,080 that at least I don't share. 189 00:09:22,080 --> 00:09:26,200 And that has been maybe present in earlier centuries. 190 00:09:26,200 --> 00:09:28,400 We don't do history like this anymore. 191 00:09:28,400 --> 00:09:31,640 But I think it's a really interesting starting point 192 00:09:31,640 --> 00:09:36,520 for discussing what I wanted to talk to you about today, 193 00:09:36,520 --> 00:09:42,120 how data about history and historical work 194 00:09:42,120 --> 00:09:45,040 as a historian-- so building historical narratives, 195 00:09:45,040 --> 00:09:47,240 making historical arguments-- 196 00:09:47,240 --> 00:09:49,560 how that works together, if it works together, 197 00:09:49,560 --> 00:09:54,560 and where it maybe also does not work very well together. 198 00:09:54,560 --> 00:09:58,640 I was hinting at the project that is behind this book. 199 00:09:58,640 --> 00:10:00,760 It's SESHAT, the Global History Data Bank. 200 00:10:00,760 --> 00:10:04,080 I don't know if anybody has had-- 201 00:10:04,080 --> 00:10:05,760 came across this before. 202 00:10:05,760 --> 00:10:08,480 I think also maybe within this particular center 203 00:10:08,480 --> 00:10:11,560 here at Leipzig University, this could be an interesting look 204 00:10:11,560 --> 00:10:13,000 at. 205 00:10:13,000 --> 00:10:17,560 The idea of this project is to collect lots of data points 206 00:10:17,560 --> 00:10:22,360 or information about history, about global history. 207 00:10:22,360 --> 00:10:29,680 And it started with building a global historical sample, 208 00:10:29,680 --> 00:10:31,960 as they call it. 209 00:10:31,960 --> 00:10:36,320 They looked for 10 world regions distributed as widely 210 00:10:36,320 --> 00:10:38,760 as possible across the Earth's surface. 211 00:10:38,760 --> 00:10:42,400 And within each of those regions designated three so-called 212 00:10:42,400 --> 00:10:47,240 natural geographical areas with discrete ecological boundaries 213 00:10:47,240 --> 00:10:50,720 in average about 10,000 square kilometers in size, 214 00:10:50,720 --> 00:10:52,160 thus creating an initial sampling 215 00:10:52,160 --> 00:10:55,320 scheme of 30 such areas around the world later extended to 35. 216 00:10:55,320 --> 00:10:57,320 This is taken from the home page in the description 217 00:10:57,320 --> 00:10:58,640 of the project. 218 00:10:58,640 --> 00:11:02,120 So this is the areas that they collect data about and data 219 00:11:02,120 --> 00:11:02,760 from. 220 00:11:02,760 --> 00:11:06,080 They do this in a very sort of a combined process 221 00:11:06,080 --> 00:11:09,360 of working together social science data, 222 00:11:09,360 --> 00:11:11,680 but also working on historical sources, 223 00:11:11,680 --> 00:11:14,360 reading, extracting that kind of information 224 00:11:14,360 --> 00:11:19,120 from what we traditionally would look at. 225 00:11:19,120 --> 00:11:21,080 And I quote the home page again on the slide. 226 00:11:21,080 --> 00:11:23,160 Most variables in SESHAD require the data 227 00:11:23,160 --> 00:11:26,920 to take the form of a number or a numerical range. 228 00:11:26,920 --> 00:11:28,840 Or they specify a feature that can be coded 229 00:11:28,840 --> 00:11:33,320 as absent, present, or unknown. 230 00:11:33,320 --> 00:11:35,120 So this is sort of the way that they 231 00:11:35,120 --> 00:11:40,360 described how the data bank, the data model actually works. 232 00:11:40,360 --> 00:11:44,840 So they usually put in numbers, and they 233 00:11:44,840 --> 00:11:49,400 can define categories as absent, present, or unknown. 234 00:11:49,400 --> 00:11:51,360 And all data are linked to scholarly sources, 235 00:11:51,360 --> 00:11:52,800 including peer-reviewed publications 236 00:11:52,800 --> 00:11:55,320 and personal communications from established authorities. 237 00:11:55,320 --> 00:11:58,080 So they also try to pick up on sort 238 00:11:58,080 --> 00:12:01,720 of the known and established historical narratives. 239 00:12:01,720 --> 00:12:03,300 This is what the home page looks like. 240 00:12:03,300 --> 00:12:06,720 So everybody, please feel invited to have a look. 241 00:12:06,720 --> 00:12:10,040 If you scroll down, these are some more of the categories 242 00:12:10,040 --> 00:12:10,840 that they look for. 243 00:12:10,840 --> 00:12:13,600 They have general variables. 244 00:12:13,600 --> 00:12:15,680 And this all-- maybe this is important-- 245 00:12:15,680 --> 00:12:18,600 it all sort of evolves around those regions. 246 00:12:18,600 --> 00:12:21,980 So they have data for a certain region. 247 00:12:21,980 --> 00:12:27,280 And everything is sort of linked to a geographical area. 248 00:12:27,280 --> 00:12:30,040 They identify 47 of such regions. 249 00:12:30,040 --> 00:12:32,040 They started with 30. 250 00:12:32,040 --> 00:12:36,360 They have now 863 polities that were present in these regions 251 00:12:36,360 --> 00:12:38,520 at one point or another. 252 00:12:38,520 --> 00:12:41,200 They look at general variables, warfare variables, 253 00:12:41,200 --> 00:12:45,120 social complexity variables, human sacrifice, 254 00:12:45,120 --> 00:12:50,680 lots of different sort of things that could be of interest. 255 00:12:50,680 --> 00:12:53,000 They themselves in this database claim 256 00:12:53,000 --> 00:12:57,800 that they don't offer a pre-constructed narrative. 257 00:12:57,800 --> 00:12:59,560 I would always set a question mark 258 00:12:59,560 --> 00:13:03,400 behind something like that, because the way that you present 259 00:13:03,400 --> 00:13:06,280 your data, the way that you collect your data, 260 00:13:06,280 --> 00:13:09,560 the way that you build the data model always already 261 00:13:09,560 --> 00:13:13,920 refers to some kind of narrative or pre-constructed 262 00:13:13,920 --> 00:13:16,120 understanding of the things that you are looking at, 263 00:13:16,120 --> 00:13:20,400 especially if you take it from text that you're reading. 264 00:13:20,400 --> 00:13:23,520 So it's not like a survey or just sort 265 00:13:23,520 --> 00:13:26,920 of counting heads of people. 266 00:13:26,920 --> 00:13:28,940 So this is the database for the book. 267 00:13:28,940 --> 00:13:30,560 Now let's have a short look at the book. 268 00:13:30,560 --> 00:13:34,080 This is the table of contents. 269 00:13:34,080 --> 00:13:36,680 And I can share this later, because you don't really 270 00:13:36,680 --> 00:13:38,720 have to read this now. 271 00:13:38,720 --> 00:13:41,520 But as you can see, they have society profiles 272 00:13:41,520 --> 00:13:44,400 starting with Egypt. 273 00:13:44,400 --> 00:13:46,320 Or they have society profiles, and they 274 00:13:46,320 --> 00:13:49,760 have sort of three categories or two categories, 275 00:13:49,760 --> 00:13:51,440 ancient and medieval. 276 00:13:51,440 --> 00:13:53,280 Because the book is trying to give us 277 00:13:53,280 --> 00:13:57,720 information about historical periods 278 00:13:57,720 --> 00:13:59,280 that we don't know so much about, 279 00:13:59,280 --> 00:14:01,880 claiming that about the modern world, we know so much 280 00:14:01,880 --> 00:14:04,400 and we have so much data that would be comparable to this. 281 00:14:04,400 --> 00:14:08,560 So essentially, one could argue that with this book 282 00:14:08,560 --> 00:14:10,600 and also with the Global History Database, 283 00:14:10,600 --> 00:14:13,960 they try to make pre-modern societies comparable 284 00:14:13,960 --> 00:14:17,160 to modern societies in a way that we 285 00:14:17,160 --> 00:14:21,680 think of society as being able to be expressed in data. 286 00:14:21,680 --> 00:14:24,000 Whether or not this is a good idea, 287 00:14:24,000 --> 00:14:25,640 we can discuss maybe later on. 288 00:14:25,640 --> 00:14:29,440 But this is sort of the general setting. 289 00:14:29,440 --> 00:14:32,520 I will show you quickly something medieval, 290 00:14:32,520 --> 00:14:35,280 because this is where I know most about. 291 00:14:35,280 --> 00:14:37,800 And we'll have a look at the rankings, 292 00:14:37,800 --> 00:14:40,480 and we'll have a very short look at also the regional adoption 293 00:14:40,480 --> 00:14:42,640 and the maps in the end. 294 00:14:42,640 --> 00:14:44,560 So this is the table of contents. 295 00:14:44,560 --> 00:14:47,120 And this is what the book looked like. 296 00:14:47,120 --> 00:14:51,480 Very authentic, because I just took pictures of the book. 297 00:14:51,480 --> 00:14:54,640 So this is what they offer us for France, 298 00:14:54,640 --> 00:14:58,160 Carolingian Kingdom, short introductory remarks. 299 00:14:58,160 --> 00:14:59,840 And then there's several categories 300 00:14:59,840 --> 00:15:01,600 where they just input data. 301 00:15:01,600 --> 00:15:03,880 And although the data bank says we only 302 00:15:03,880 --> 00:15:10,520 have categories or labels that we fill with the numbers, 303 00:15:10,520 --> 00:15:12,760 there's some text in here. 304 00:15:12,760 --> 00:15:18,960 So it starts with the duration, 752 to 987 CE. 305 00:15:18,960 --> 00:15:20,920 Being a medieval historian myself, 306 00:15:20,920 --> 00:15:22,960 this is debatable already. 307 00:15:22,960 --> 00:15:26,920 This is like the first entry, and we would say, yeah, OK. 308 00:15:26,920 --> 00:15:29,680 Why this, why not other? 309 00:15:29,680 --> 00:15:31,880 I mean, I probably don't have to explain this 310 00:15:31,880 --> 00:15:34,640 to the people sitting here, but it's debatable 311 00:15:34,640 --> 00:15:37,080 whether this is the correct date. 312 00:15:37,080 --> 00:15:38,320 Then they have the total area. 313 00:15:38,320 --> 00:15:41,200 Again, at what point in time, what is the extension, 314 00:15:41,200 --> 00:15:42,320 how did this change? 315 00:15:42,320 --> 00:15:46,360 So as you can see, being a medieval historian myself, 316 00:15:46,360 --> 00:15:50,600 I immediately start to question not the facticity, 317 00:15:50,600 --> 00:15:53,560 but maybe sort of the point in time 318 00:15:53,560 --> 00:15:55,320 that this data point represents. 319 00:15:55,320 --> 00:15:57,240 So I want to know, why did you decide 320 00:15:57,240 --> 00:15:58,680 to say this is the duration? 321 00:15:58,680 --> 00:16:00,440 I want to decide why, or I want to know 322 00:16:00,440 --> 00:16:05,480 why they decided this is the area that it covered. 323 00:16:05,480 --> 00:16:08,760 Then we have institutions, legal code absent. 324 00:16:08,760 --> 00:16:14,120 The Monumenta Germania Historia in Munich, 325 00:16:14,120 --> 00:16:16,400 which is an institution that at least medieval 326 00:16:16,400 --> 00:16:18,800 or pre-modern historians know very well, 327 00:16:18,800 --> 00:16:23,720 are just on the way of doing a re-edition of the Carolingian 328 00:16:23,720 --> 00:16:26,200 legal codes, the Capitularia. 329 00:16:26,200 --> 00:16:28,880 So again, I'm not so sure. 330 00:16:28,880 --> 00:16:33,440 I don't want to bash this, but just to make a point 331 00:16:33,440 --> 00:16:37,720 that they're trying to data find something that might not 332 00:16:37,720 --> 00:16:41,960 fit this general idea of how we deal with data points 333 00:16:41,960 --> 00:16:43,520 that we might have today. 334 00:16:43,520 --> 00:16:49,840 And if one of the arguments is to make a book like this 335 00:16:49,840 --> 00:16:52,600 in order to make pre-modern societies 336 00:16:52,600 --> 00:16:56,440 comparable to modern societies, is that the right way 337 00:16:56,440 --> 00:16:57,040 to compare? 338 00:16:57,040 --> 00:16:59,200 Is it even the right way to describe modern societies 339 00:16:59,200 --> 00:17:02,320 if we look at modern societies through a data-fied lens 340 00:17:02,320 --> 00:17:05,080 or a data perspective? 341 00:17:05,080 --> 00:17:07,920 Just because it's so much fun, we're going to look at this. 342 00:17:07,920 --> 00:17:11,360 10 largest field armies, ancient, medieval, early 343 00:17:11,360 --> 00:17:15,400 modern, in the entire pre-industrial area. 344 00:17:15,400 --> 00:17:17,640 This is actually interesting because I have no idea 345 00:17:17,640 --> 00:17:21,160 about military history, so this is blank for me. 346 00:17:21,160 --> 00:17:26,440 I cannot argue as much as I could with the previous slide. 347 00:17:26,440 --> 00:17:30,240 But what is interesting is that this is actually not-- 348 00:17:30,240 --> 00:17:31,760 for my students, I would use this 349 00:17:31,760 --> 00:17:34,040 to show them that they need to get 350 00:17:34,040 --> 00:17:36,160 rid of their European focus. 351 00:17:36,160 --> 00:17:39,800 Because medieval history in German universities 352 00:17:39,800 --> 00:17:43,400 is mostly Latin Europe. 353 00:17:43,400 --> 00:17:47,400 And you can see immediately that at least withstanding armies, 354 00:17:47,400 --> 00:17:53,160 there is not so many European polities in the list. 355 00:17:53,160 --> 00:17:55,200 Of course, here's the Romans. 356 00:17:55,200 --> 00:18:01,360 And there's lots of East Asian areas on the top. 357 00:18:01,360 --> 00:18:03,400 This changes. 358 00:18:03,400 --> 00:18:05,520 So here's the Ottoman Empire. 359 00:18:05,520 --> 00:18:10,160 And up there, like for the early modern period, 360 00:18:10,160 --> 00:18:12,720 we have the Bourbons on top. 361 00:18:12,720 --> 00:18:14,680 Interesting. 362 00:18:14,680 --> 00:18:18,520 They have a larger standing army than some of the regions 363 00:18:18,520 --> 00:18:19,520 that were bigger before. 364 00:18:19,520 --> 00:18:26,840 So I forgot to show you the map. 365 00:18:26,840 --> 00:18:29,200 Maybe I can do this over drinks afterwards. 366 00:18:29,200 --> 00:18:30,840 Sorry, it's gone. 367 00:18:30,840 --> 00:18:36,880 So Turchin argues that we understand our own world 368 00:18:36,880 --> 00:18:38,640 through data, and this is an argument 369 00:18:38,640 --> 00:18:41,400 that he makes for his contemporary world, 370 00:18:41,400 --> 00:18:43,120 vast arrays of statistics that reveal 371 00:18:43,120 --> 00:18:45,360 the workings of societies. 372 00:18:45,360 --> 00:18:48,680 I would then like to ask him, what kind of data do you get? 373 00:18:48,680 --> 00:18:49,600 How do you collect it? 374 00:18:49,600 --> 00:18:51,600 I made this excursion into the data bank 375 00:18:51,600 --> 00:18:56,760 to give you an idea of how this generally works. 376 00:18:56,760 --> 00:19:01,080 The statement makes it sound as though dealing with data 377 00:19:01,080 --> 00:19:04,160 and doing statistics is a very contemporary thing, 378 00:19:04,160 --> 00:19:06,840 or at least a very modern thing. 379 00:19:06,840 --> 00:19:09,320 Again, maybe this is a void argument within this audience, 380 00:19:09,320 --> 00:19:11,400 but obviously-- or of course, it is not. 381 00:19:11,400 --> 00:19:13,760 So there have been lots of historical data practices 382 00:19:13,760 --> 00:19:14,800 before. 383 00:19:14,800 --> 00:19:17,800 And even Amina Say, who some of you might have read, 384 00:19:17,800 --> 00:19:19,840 he published a book called Musta in German, 385 00:19:19,840 --> 00:19:22,760 now translated into English as Patterns, 386 00:19:22,760 --> 00:19:24,520 a theory of the digitized society, 387 00:19:24,520 --> 00:19:27,920 argues that digitization, and with it 388 00:19:27,920 --> 00:19:31,200 the understanding of data that is most common today, 389 00:19:31,200 --> 00:19:37,720 is not something that kind of started when we-- 390 00:19:37,720 --> 00:19:40,840 at a point in time where the technology was far ahead 391 00:19:40,840 --> 00:19:43,720 enough to represent what we understand when 392 00:19:43,720 --> 00:19:45,880 we talk about digitization. 393 00:19:45,880 --> 00:19:48,840 It's basically ones and zeros. 394 00:19:48,840 --> 00:19:52,240 So it's trying to find something that is quantifiable, 395 00:19:52,240 --> 00:19:54,040 that I can count. 396 00:19:54,040 --> 00:19:57,840 Digitization comes from digitus, taking your fingers. 397 00:19:57,840 --> 00:20:00,120 So it's basically counting, right? 398 00:20:00,120 --> 00:20:02,320 And these kinds of practices were in the world 399 00:20:02,320 --> 00:20:06,040 long before what we today understand as digitization, 400 00:20:06,040 --> 00:20:09,520 and also long before that what we talk about-- 401 00:20:09,520 --> 00:20:13,280 what we think about when we talk about data or statistics. 402 00:20:13,280 --> 00:20:15,640 I think this is important to understand 403 00:20:15,640 --> 00:20:21,400 when we try to evaluate how or which role quantifiable facts 404 00:20:21,400 --> 00:20:24,920 or data plays in constructing historical narratives, which 405 00:20:24,920 --> 00:20:30,240 is where I will be going by the end of this talk, 406 00:20:30,240 --> 00:20:33,280 that it's not something that is here just because we 407 00:20:33,280 --> 00:20:35,520 have computers now. 408 00:20:35,520 --> 00:20:37,680 And we've been talking about this earlier. 409 00:20:37,680 --> 00:20:40,040 There's a long history in historiography 410 00:20:40,040 --> 00:20:42,520 also with social history and economic history 411 00:20:42,520 --> 00:20:45,800 of the '70s and '80s where people counted things, 412 00:20:45,800 --> 00:20:47,200 historical demography. 413 00:20:47,200 --> 00:20:50,960 There's lots of different areas that 414 00:20:50,960 --> 00:20:53,200 are part of what we understand as history 415 00:20:53,200 --> 00:20:56,320 as a discipline where quantification is very present. 416 00:20:56,320 --> 00:20:59,000 We've had these discussions whether numbers tell stories 417 00:20:59,000 --> 00:21:01,280 or not decades ago. 418 00:21:01,280 --> 00:21:02,400 We have them now again. 419 00:21:02,400 --> 00:21:06,200 And nobody talks about why don't we relate the two discussions 420 00:21:06,200 --> 00:21:07,760 to each other. 421 00:21:07,760 --> 00:21:11,720 So why do I still have to justify 422 00:21:11,720 --> 00:21:14,680 that I work with quantification-- 423 00:21:14,680 --> 00:21:17,560 setting, machine learning, and all these fancy stuff aside? 424 00:21:17,560 --> 00:21:21,080 Basically, it's quantification when 425 00:21:21,080 --> 00:21:23,720 we had this discussion like 70 years ago 426 00:21:23,720 --> 00:21:25,780 and again and again and again. 427 00:21:25,780 --> 00:21:27,320 So also for history as a discipline, 428 00:21:27,320 --> 00:21:29,200 this is an interesting observation. 429 00:21:29,200 --> 00:21:31,200 I don't quite know what to do with it yet. 430 00:21:31,200 --> 00:21:33,400 But I think there's also a narrative there 431 00:21:33,400 --> 00:21:35,520 that we should have a look at. 432 00:21:35,520 --> 00:21:37,840 So if we follow Amin Ase and not only him, 433 00:21:37,840 --> 00:21:40,640 he argues that digitalization is not only 434 00:21:40,640 --> 00:21:43,040 a contemporary phenomenon but generally helps societies 435 00:21:43,040 --> 00:21:45,000 to deal with and reduce complexity 436 00:21:45,000 --> 00:21:47,920 by using coded numbers to process information. 437 00:21:47,920 --> 00:21:50,060 Make something quantifiable. 438 00:21:50,060 --> 00:21:51,720 Make it countable. 439 00:21:51,720 --> 00:21:56,720 This makes it easier to get an overview, to get control, 440 00:21:56,720 --> 00:22:02,760 or apparently get control over complex situations. 441 00:22:02,760 --> 00:22:05,200 Which are the data practices that we know historically 442 00:22:05,200 --> 00:22:07,280 and how do we use data practices in history? 443 00:22:07,280 --> 00:22:09,480 I've started to kind of get into that question 444 00:22:09,480 --> 00:22:11,440 with my earlier remarks. 445 00:22:11,440 --> 00:22:14,560 Just a few very random examples. 446 00:22:14,560 --> 00:22:17,940 This is a contemporary map from 1903. 447 00:22:17,940 --> 00:22:21,920 I did a seminar, bachelor's seminar, 448 00:22:21,920 --> 00:22:26,560 at Bielefeld University where we worked on election data 449 00:22:26,560 --> 00:22:31,080 from previous German elections. 450 00:22:31,080 --> 00:22:35,060 And the students were really surprised that in 1903, we 451 00:22:35,060 --> 00:22:38,240 already had all these kinds of diagrams. 452 00:22:38,240 --> 00:22:39,440 They have bar diagrams. 453 00:22:39,440 --> 00:22:41,440 They have pie-- they have it all. 454 00:22:41,440 --> 00:22:43,720 So why is this? 455 00:22:43,720 --> 00:22:45,720 I mean, this is 120 years ago. 456 00:22:45,720 --> 00:22:49,080 So there was some irritation, which 457 00:22:49,080 --> 00:22:50,920 I think is good to start to get them 458 00:22:50,920 --> 00:22:53,980 into questioning certain things. 459 00:22:53,980 --> 00:22:55,360 So the way that we deal with data 460 00:22:55,360 --> 00:22:58,320 and also visualizations that are so natural to us 461 00:22:58,320 --> 00:23:00,600 have been there for quite some time. 462 00:23:00,600 --> 00:23:05,160 Another example, all the way back at the beginning 463 00:23:05,160 --> 00:23:07,560 of the 12th century, "Doom's Day Book." 464 00:23:07,560 --> 00:23:11,000 I think most of you have already maybe heard of this. 465 00:23:11,000 --> 00:23:15,420 This is the first survey of a register 466 00:23:15,420 --> 00:23:24,820 of the British population, stating also the land 467 00:23:24,820 --> 00:23:26,540 that belonged to certain people. 468 00:23:26,540 --> 00:23:29,380 And this is a register that is still more or less valid today 469 00:23:29,380 --> 00:23:32,880 because you can go back all the way to the time of William 470 00:23:32,880 --> 00:23:37,100 the Conqueror just after 1066 when he started to record this. 471 00:23:37,100 --> 00:23:42,380 I think the first survey of the "Doom's Day Book" is 1080, 472 00:23:42,380 --> 00:23:45,380 so at the late 11th century. 473 00:23:45,380 --> 00:23:48,840 So counting people, trying to pinpoint areas 474 00:23:48,840 --> 00:23:50,700 in which people live, and thinking 475 00:23:50,700 --> 00:23:53,060 about who is in charge of which area, 476 00:23:53,060 --> 00:23:57,440 how does the politics kind of work 477 00:23:57,440 --> 00:24:00,140 on the basis of data and statistics is nothing 478 00:24:00,140 --> 00:24:03,380 that we haven't seen before. 479 00:24:03,380 --> 00:24:05,300 It has a long tradition. 480 00:24:05,300 --> 00:24:07,980 So what about the way that we deal with this today? 481 00:24:07,980 --> 00:24:09,860 And how does that-- 482 00:24:09,860 --> 00:24:12,860 maybe sort of reopening the discussion 483 00:24:12,860 --> 00:24:16,420 about quantification in history, or quantified sources, 484 00:24:16,420 --> 00:24:21,180 or like tables and numbers as sources in history-- 485 00:24:21,180 --> 00:24:23,300 what does that have to do with digital history? 486 00:24:23,300 --> 00:24:27,380 And why do I think that we still need to think about this? 487 00:24:27,380 --> 00:24:30,280 Here's another example. 488 00:24:30,280 --> 00:24:33,740 And this is important because I'm getting back to this 489 00:24:33,740 --> 00:24:36,560 when I talk about the VR project. 490 00:24:36,560 --> 00:24:41,720 This is one page within a big book. 491 00:24:41,720 --> 00:24:45,840 It was first published in 1952 by Arno Peters and his wife. 492 00:24:45,840 --> 00:24:49,040 Arno Peters was a historian and geographer, 493 00:24:49,040 --> 00:24:51,000 and he collected lots of information 494 00:24:51,000 --> 00:24:55,480 again about historical events in different world regions, 495 00:24:55,480 --> 00:24:57,440 and he published a book about it. 496 00:24:57,440 --> 00:24:59,140 And it's a big book, and when you open it, 497 00:24:59,140 --> 00:25:00,800 you always have one of these big pages, 498 00:25:00,800 --> 00:25:05,260 as you can see on the screen, in front of you. 499 00:25:05,260 --> 00:25:10,080 And this always covers 100 years from here to here. 500 00:25:10,080 --> 00:25:11,900 The middle section gives you people 501 00:25:11,900 --> 00:25:16,940 who lived within the century according to their birth dates 502 00:25:16,940 --> 00:25:22,260 and death dates. 503 00:25:22,260 --> 00:25:25,700 And also you have certain events that are 504 00:25:25,700 --> 00:25:27,680 categorized in different colors. 505 00:25:27,680 --> 00:25:31,300 This down here is wars and revolutions. 506 00:25:31,300 --> 00:25:36,660 Up there is inventions and economics, and also something 507 00:25:36,660 --> 00:25:38,460 that we would call maybe history of ideas 508 00:25:38,460 --> 00:25:40,460 or intellectual history. 509 00:25:40,460 --> 00:25:45,020 This is a practice where he tried to also kind of open 510 00:25:45,020 --> 00:25:50,860 this idea of collecting data about, in this case, 511 00:25:50,860 --> 00:25:55,580 historical societies, and not only having a European view, 512 00:25:55,580 --> 00:25:57,940 but also making this-- 513 00:25:57,940 --> 00:26:01,340 I'm always sort of hesitant to use the term global, 514 00:26:01,340 --> 00:26:03,100 because it's much more complex, I think, 515 00:26:03,100 --> 00:26:05,580 than what I usually say with it. 516 00:26:05,580 --> 00:26:08,220 But it's trying to sort of open a perspective that 517 00:26:08,220 --> 00:26:10,860 goes beyond sort of the European perspective 518 00:26:10,860 --> 00:26:14,240 and give you information about historical events, whatever 519 00:26:14,240 --> 00:26:19,900 that is, and people that were present in a certain century. 520 00:26:19,900 --> 00:26:23,300 This is a picture of the data collected 521 00:26:23,300 --> 00:26:26,020 that is in the background. 522 00:26:26,020 --> 00:26:34,060 The [SPEAKING GERMAN] contains 12,162 events. 523 00:26:34,060 --> 00:26:37,260 Some of them are geo-referenced, but not all of them. 524 00:26:37,260 --> 00:26:39,260 There's information about persons. 525 00:26:39,260 --> 00:26:43,940 There is a keyword register that is something like a dictionary, 526 00:26:43,940 --> 00:26:45,920 but there's no real understanding 527 00:26:45,920 --> 00:26:50,200 of why these words are the important ones. 528 00:26:50,200 --> 00:26:52,800 There are also connections within the data set, 529 00:26:52,800 --> 00:26:58,680 so which of the entries is sort of related to another one. 530 00:26:58,680 --> 00:27:01,880 And these are the entry cards that the people 531 00:27:01,880 --> 00:27:06,320 made in the '50s and '60s when the book was sort of put 532 00:27:06,320 --> 00:27:07,240 together. 533 00:27:07,240 --> 00:27:09,240 And you can see this is all typewriter. 534 00:27:09,240 --> 00:27:11,760 This is all German, unfortunately. 535 00:27:11,760 --> 00:27:13,340 And then you have different categories. 536 00:27:13,340 --> 00:27:15,220 You have some handwriting for the datings 537 00:27:15,220 --> 00:27:16,340 and all that kind of thing. 538 00:27:16,340 --> 00:27:20,820 So you can think of a big [SPEAKING GERMAN] 539 00:27:20,820 --> 00:27:23,000 that they had in order to make this work. 540 00:27:23,000 --> 00:27:25,340 And of course, this has been digitized now. 541 00:27:25,340 --> 00:27:32,580 And although I say this myself, it basically looks the same. 542 00:27:32,580 --> 00:27:40,940 Same effort, same data set, same representation. 543 00:27:40,940 --> 00:27:43,460 So when I first opened this, I was a bit disappointed, 544 00:27:43,460 --> 00:27:45,480 because I thought, OK, you have the big data set, 545 00:27:45,480 --> 00:27:46,880 so what do you do with it? 546 00:27:46,880 --> 00:27:47,960 First of all, and this is something 547 00:27:47,960 --> 00:27:50,120 that happens in digital humanities and digital history 548 00:27:50,120 --> 00:27:53,120 most of the time, it can be read as something that 549 00:27:53,120 --> 00:27:54,680 is called a proof of concept. 550 00:27:54,680 --> 00:27:58,100 So I show you that my digital version can do the same thing 551 00:27:58,100 --> 00:28:00,920 that the paper version can. 552 00:28:00,920 --> 00:28:03,840 And once I've established that, I can take the next step 553 00:28:03,840 --> 00:28:06,020 and show you what the digital version can, 554 00:28:06,020 --> 00:28:08,800 what the paper version cannot, which is basically, 555 00:28:08,800 --> 00:28:11,600 in this case, rearranging information 556 00:28:11,600 --> 00:28:15,320 and digging deep into the connections that 557 00:28:15,320 --> 00:28:17,500 are present in the database. 558 00:28:17,500 --> 00:28:19,060 So again, I explained this before. 559 00:28:19,060 --> 00:28:20,500 So here's the people. 560 00:28:20,500 --> 00:28:23,060 Then there's events in these different categories. 561 00:28:23,060 --> 00:28:24,780 You have some pictures. 562 00:28:24,780 --> 00:28:26,220 You can now-- this is interactive, 563 00:28:26,220 --> 00:28:28,140 so you can click on it, and then you 564 00:28:28,140 --> 00:28:30,380 can see the information that was entered. 565 00:28:30,380 --> 00:28:34,300 So this is the first upgrade that the digital version 566 00:28:34,300 --> 00:28:37,140 has over the paper version. 567 00:28:37,140 --> 00:28:39,680 And this is-- oh, sorry, I have this in bigger, 568 00:28:39,680 --> 00:28:42,140 so you can have a closer look. 569 00:28:42,140 --> 00:28:44,780 This is the 14th century. 570 00:28:44,780 --> 00:28:47,120 Again, there's the medieval training 571 00:28:47,120 --> 00:28:48,580 coming through at some points. 572 00:28:48,580 --> 00:28:56,060 This is more interesting because this is a visualization 573 00:28:56,060 --> 00:29:00,500 that the digital version of the Zunkornapturshir world history 574 00:29:00,500 --> 00:29:04,340 offers us that gives us a different view on the data 575 00:29:04,340 --> 00:29:06,140 sets that they have. 576 00:29:06,140 --> 00:29:10,220 This, going back one, is still a representation 577 00:29:10,220 --> 00:29:12,520 of history in a straight line. 578 00:29:12,520 --> 00:29:15,600 This is a very traditional, very linear, very kind 579 00:29:15,600 --> 00:29:19,760 of [SPEAKING GERMAN] history of events kind of approach. 580 00:29:19,760 --> 00:29:23,160 Also, important people, most of them are male. 581 00:29:23,160 --> 00:29:25,800 I'm just mentioning this because it's part of the truth. 582 00:29:25,800 --> 00:29:32,960 But now, you can kind of rearrange the data points 583 00:29:32,960 --> 00:29:36,220 and play around with the perspective also on history 584 00:29:36,220 --> 00:29:38,880 as something that happens in time. 585 00:29:38,880 --> 00:29:40,340 What this visualization does is you 586 00:29:40,340 --> 00:29:41,960 have one of the events in the middle. 587 00:29:41,960 --> 00:29:43,620 In this case, this is the same event 588 00:29:43,620 --> 00:29:47,420 that I had the information card about. 589 00:29:47,420 --> 00:29:50,580 It's 1331, the invention of canons, 590 00:29:50,580 --> 00:29:55,500 or the first use of canons in Priole in Italy. 591 00:29:55,500 --> 00:29:58,340 And then you can see other events or entries 592 00:29:58,340 --> 00:30:01,580 in the database that are related to the one in the middle. 593 00:30:01,580 --> 00:30:03,420 It's not linear anymore. 594 00:30:03,420 --> 00:30:07,340 It's more layered, right? 595 00:30:07,340 --> 00:30:10,900 So I have sort of time layers that surround 596 00:30:10,900 --> 00:30:12,860 the event in the middle. 597 00:30:12,860 --> 00:30:15,680 So the further away on the layers, 598 00:30:15,680 --> 00:30:18,120 the further away in time. 599 00:30:18,120 --> 00:30:20,280 And this can be either before or after. 600 00:30:20,280 --> 00:30:22,880 So it's not a linear-- 601 00:30:22,880 --> 00:30:24,740 so this is not early, and this is late. 602 00:30:24,740 --> 00:30:28,120 It's just sort of the absolute distance. 603 00:30:28,120 --> 00:30:31,960 And the size of the nodes represents 604 00:30:31,960 --> 00:30:36,680 the semantic relation or the closeness of the events 605 00:30:36,680 --> 00:30:42,600 to each other in a more hermeneutic way. 606 00:30:42,600 --> 00:30:44,440 Interesting is we try to understand 607 00:30:44,440 --> 00:30:48,040 how this works computationally, and it's just random. 608 00:30:48,040 --> 00:30:50,440 It's really random. 609 00:30:50,440 --> 00:30:54,240 But the idea is that the bigger the node, the closer 610 00:30:54,240 --> 00:30:57,080 the events are to each other, like semantically 611 00:30:57,080 --> 00:30:58,980 or hermeneutically speaking. 612 00:30:58,980 --> 00:31:01,200 Just one example down here is the Hussites 613 00:31:01,200 --> 00:31:03,320 who use the haubitsa for the first time. 614 00:31:03,320 --> 00:31:06,920 So this, I get, is quite similar as a event. 615 00:31:06,920 --> 00:31:08,760 So they developed a weapon. 616 00:31:08,760 --> 00:31:10,280 They're using it for the first time. 617 00:31:10,280 --> 00:31:13,480 OK, right, so that those are more or less related. 618 00:31:13,480 --> 00:31:13,960 I get that. 619 00:31:13,960 --> 00:31:25,040 So I think this shows that data sets and the way-- 620 00:31:25,040 --> 00:31:29,280 or rather, the way that we use, in this case, visualizations 621 00:31:29,280 --> 00:31:34,000 and representations of data sets that already exist 622 00:31:34,000 --> 00:31:38,440 or that we kind of also generate changes our perspective 623 00:31:38,440 --> 00:31:40,880 on the kinds of stories that we tell. 624 00:31:40,880 --> 00:31:43,680 If I use this, I'm not going to start linear. 625 00:31:43,680 --> 00:31:46,240 I'm going to start in a relational way. 626 00:31:46,240 --> 00:31:48,660 So this is a representation of a relational way 627 00:31:48,660 --> 00:31:52,280 to tell historical or build historical narratives, 628 00:31:52,280 --> 00:31:55,440 rather than just having it in a linear way. 629 00:31:55,440 --> 00:31:59,820 And this is, I think, one takeaway for me. 630 00:31:59,820 --> 00:32:01,240 Using these kinds of visualizations 631 00:32:01,240 --> 00:32:03,600 and playing around and exploring different kinds 632 00:32:03,600 --> 00:32:06,300 of visualizations for the same data sets 633 00:32:06,300 --> 00:32:09,120 helps me to sort of open up different perspectives 634 00:32:09,120 --> 00:32:12,000 and maybe gain more than just one perspective 635 00:32:12,000 --> 00:32:15,400 on a certain historical topic. 636 00:32:15,400 --> 00:32:19,200 So I already started with my last example 637 00:32:19,200 --> 00:32:20,660 to go into the complex of questions 638 00:32:20,660 --> 00:32:22,400 that I want to talk about now. 639 00:32:22,400 --> 00:32:25,060 Is it really data versus stories? 640 00:32:25,060 --> 00:32:29,620 Or how do stories maybe also consist of data? 641 00:32:29,620 --> 00:32:34,500 Or how do certain data arrays or arrangements, data sets, 642 00:32:34,500 --> 00:32:38,700 visualizations, guide the way that I tell stories? 643 00:32:38,700 --> 00:32:40,320 I could now also give you a lecture 644 00:32:40,320 --> 00:32:42,780 on what historians usually do, but I'm not going to do that. 645 00:32:42,780 --> 00:32:44,720 Because I see some faces here that 646 00:32:44,720 --> 00:32:46,860 have done this for generations of students already. 647 00:32:46,860 --> 00:32:50,300 So we're not going to go down there. 648 00:32:50,300 --> 00:32:53,160 But I think it's important to reflect-- 649 00:32:53,160 --> 00:32:55,920 and this is why I made so clear that I'm a digital historian. 650 00:32:55,920 --> 00:32:58,940 Not only a humanist, I'm a historian first. 651 00:32:58,940 --> 00:33:01,680 Because I have a certain approach and a certain kind 652 00:33:01,680 --> 00:33:04,080 that I ask my questions when I look at data sets. 653 00:33:04,080 --> 00:33:05,840 And I think this is important, because this 654 00:33:05,840 --> 00:33:10,080 is different to my computational literary scholars, colleagues. 655 00:33:10,080 --> 00:33:11,600 They work with data sets as well. 656 00:33:11,600 --> 00:33:12,960 They work with the same methods, but they 657 00:33:12,960 --> 00:33:14,360 have different sets of questions. 658 00:33:14,360 --> 00:33:16,960 And they want to sort of tell other stories 659 00:33:16,960 --> 00:33:21,840 if we phrase it like that. 660 00:33:21,840 --> 00:33:23,900 OK, so what we do as historians is-- 661 00:33:23,900 --> 00:33:27,380 and I'm going to do this in just one sentence-- 662 00:33:27,380 --> 00:33:32,300 we try to build plausible narratives about the past. 663 00:33:32,300 --> 00:33:35,060 It's important that they are plausible. 664 00:33:35,060 --> 00:33:37,440 They are not facts in and of itself. 665 00:33:37,440 --> 00:33:38,480 They consist of facts. 666 00:33:38,480 --> 00:33:39,960 They consist of data. 667 00:33:39,960 --> 00:33:43,160 But there's not just one story about something. 668 00:33:43,160 --> 00:33:45,640 So it's important to me that we try-- 669 00:33:45,640 --> 00:33:47,420 that we sort of focus on the plausible. 670 00:33:47,420 --> 00:33:48,920 Because I think that for historians 671 00:33:48,920 --> 00:33:51,520 and for other hermeneutic disciplines, 672 00:33:51,520 --> 00:33:54,960 some kind of vagueness and some kind 673 00:33:54,960 --> 00:33:56,920 of sort of interpretation and leeway 674 00:33:56,920 --> 00:33:59,560 is really important to the way that we work. 675 00:33:59,560 --> 00:34:01,440 We always try to differentiate things. 676 00:34:01,440 --> 00:34:03,320 It's just not one way. 677 00:34:03,320 --> 00:34:08,120 There's the most plausible way, but it's not the only way. 678 00:34:08,120 --> 00:34:10,200 And this is something that is really complicated 679 00:34:10,200 --> 00:34:14,920 to match to data sets, because data sets usually are one way. 680 00:34:14,920 --> 00:34:16,360 It's one data point. 681 00:34:16,360 --> 00:34:17,360 It is one category. 682 00:34:17,360 --> 00:34:21,600 They might have different properties. 683 00:34:21,600 --> 00:34:24,040 I can try to kind of build a complex data model where 684 00:34:24,040 --> 00:34:27,120 I have lots of different layers in my annotation scheme, 685 00:34:27,120 --> 00:34:31,200 and it goes like a tree in sort of different directions. 686 00:34:31,200 --> 00:34:33,240 But basically, I always have to make a decision 687 00:34:33,240 --> 00:34:36,040 if it is yes or no in the end. 688 00:34:36,040 --> 00:34:39,520 And this is something that can be very painful for historians 689 00:34:39,520 --> 00:34:42,760 if they have to, say, make that particular decision 690 00:34:42,760 --> 00:34:46,920 in a historical case. 691 00:34:46,920 --> 00:34:51,280 And because it is so difficult and sometimes painful, 692 00:34:51,280 --> 00:34:57,080 it is important to dwell on this maybe just for three seconds 693 00:34:57,080 --> 00:34:58,960 longer. 694 00:34:58,960 --> 00:35:02,040 Historical narratives might be based on data 695 00:35:02,040 --> 00:35:03,800 or are basically always based on data 696 00:35:03,800 --> 00:35:06,200 if you think of all kinds of information 697 00:35:06,200 --> 00:35:09,440 can be data that represent a certain perspective. 698 00:35:09,440 --> 00:35:11,200 It's always made. 699 00:35:11,200 --> 00:35:15,320 Johanna Drucker, one of, I think, the most interesting 700 00:35:15,320 --> 00:35:19,200 and also sort of already long-term digital humanists 701 00:35:19,200 --> 00:35:22,440 in the United States, in one of her publications, 702 00:35:22,440 --> 00:35:25,680 she claims that we should stop calling it data 703 00:35:25,680 --> 00:35:30,000 and start calling it captor from Latin "catching," 704 00:35:30,000 --> 00:35:31,760 because data is always a given. 705 00:35:31,760 --> 00:35:34,040 That's sort of the literal translation. 706 00:35:34,040 --> 00:35:36,560 And she says we should rather call it captor because it 707 00:35:36,560 --> 00:35:38,480 is something that we take and that we form 708 00:35:38,480 --> 00:35:42,280 and that we sort of decide what it should look like. 709 00:35:42,280 --> 00:35:43,520 And I think this is important. 710 00:35:43,520 --> 00:35:45,480 Every database has a perspective. 711 00:35:45,480 --> 00:35:48,720 It's not just there because it is there. 712 00:35:48,720 --> 00:35:52,200 Even the decision that the people of the Global History 713 00:35:52,200 --> 00:35:54,080 Data Bank made-- 714 00:35:54,080 --> 00:35:56,240 do I calculate heads? 715 00:35:56,240 --> 00:35:59,560 How big is the area that I define as the area? 716 00:35:59,560 --> 00:36:02,480 This is all part of that particular perspective. 717 00:36:02,480 --> 00:36:04,940 And this is true for narratives as well as for data sets. 718 00:36:04,940 --> 00:36:07,160 There's no difference. 719 00:36:07,160 --> 00:36:10,680 Maybe narratives can be more elaborate about this. 720 00:36:10,680 --> 00:36:12,760 Data sets need a really good documentation 721 00:36:12,760 --> 00:36:16,920 that people can follow in order to understand this. 722 00:36:16,920 --> 00:36:20,200 If this is true, that they are all 723 00:36:20,200 --> 00:36:21,620 dependent on a certain perspective, 724 00:36:21,620 --> 00:36:23,900 they are oriented in knowledge orders and models. 725 00:36:23,900 --> 00:36:25,280 So there's a certain understanding. 726 00:36:25,280 --> 00:36:29,040 I think I started this talk also with making the point 727 00:36:29,040 --> 00:36:33,480 that it's always also about sort of a hidden understanding 728 00:36:33,480 --> 00:36:34,600 of something. 729 00:36:34,600 --> 00:36:37,880 If I make a database, I have presumptions. 730 00:36:37,880 --> 00:36:40,800 And I think that what historians can learn 731 00:36:40,800 --> 00:36:43,400 in the process of creating databases that 732 00:36:43,400 --> 00:36:45,440 need a good documentation is to be 733 00:36:45,440 --> 00:36:49,760 more precise about what we think that these presumptions are. 734 00:36:49,760 --> 00:36:51,820 Because we usually work in footnotes, 735 00:36:51,820 --> 00:36:54,520 which are also very differentiating. 736 00:36:54,520 --> 00:36:58,040 Maybe there's still interpretation and arguments. 737 00:36:58,040 --> 00:37:01,400 And I think we can learn from documentation to process 738 00:37:01,400 --> 00:37:06,000 the research decisions that we make, 739 00:37:06,000 --> 00:37:08,160 document them more closely. 740 00:37:08,160 --> 00:37:12,840 And I think this is like another takeaway if you work with data. 741 00:37:12,840 --> 00:37:15,600 Not to say digitally, because we all work digitally, 742 00:37:15,600 --> 00:37:18,440 but to work with data. 743 00:37:18,440 --> 00:37:21,640 I'm coming back to that perspective maybe in a bit. 744 00:37:21,640 --> 00:37:25,840 The last point on this slide is something 745 00:37:25,840 --> 00:37:28,520 that I will come to later when I give you 746 00:37:28,520 --> 00:37:31,000 some examples of visualizations. 747 00:37:31,000 --> 00:37:37,840 I think that one point or one element of also 748 00:37:37,840 --> 00:37:44,240 the digital version of the PTAS helps to understand this-- 749 00:37:44,240 --> 00:37:46,440 let's call it perspectiveness or being 750 00:37:46,440 --> 00:37:48,280 bound to a certain perspective. 751 00:37:48,280 --> 00:37:49,440 This is interactive. 752 00:37:49,440 --> 00:37:53,680 I can change the perspective, and I see the image changing. 753 00:37:53,680 --> 00:37:57,960 This is really important to give users this understanding of, 754 00:37:57,960 --> 00:38:01,040 if I stand on another point in the data set, 755 00:38:01,040 --> 00:38:03,440 I see other things around me. 756 00:38:03,440 --> 00:38:06,120 And this is basically the starting point 757 00:38:06,120 --> 00:38:10,320 for the virtual reality project, which basically works like this. 758 00:38:10,320 --> 00:38:14,320 I can stand within a knowledge graph on a certain node, 759 00:38:14,320 --> 00:38:15,400 and then I can move. 760 00:38:15,400 --> 00:38:18,400 And then the whole thing changes around me 761 00:38:18,400 --> 00:38:23,240 so that I understand I can only see in this cone in front 762 00:38:23,240 --> 00:38:23,720 of me. 763 00:38:23,720 --> 00:38:25,200 Maybe I have to turn around to see something 764 00:38:25,200 --> 00:38:26,720 that is almost as interesting. 765 00:38:26,720 --> 00:38:29,000 But I've never looked at this part of the knowledge graph 766 00:38:29,000 --> 00:38:30,320 before. 767 00:38:30,320 --> 00:38:34,680 So narratives are one perspective. 768 00:38:34,680 --> 00:38:37,160 They might be multilayered and have 769 00:38:37,160 --> 00:38:39,240 integrated more than one perspective. 770 00:38:39,240 --> 00:38:41,400 But in the end, if you tell the story, 771 00:38:41,400 --> 00:38:45,080 it comes to one certain set of perspectives. 772 00:38:45,080 --> 00:38:47,880 Same is true for data sets. 773 00:38:47,880 --> 00:38:50,880 Can we then use data sets maybe also in the same way 774 00:38:50,880 --> 00:38:53,320 and make them tell stories? 775 00:38:53,320 --> 00:38:55,280 Or how do we-- and this is one of the questions 776 00:38:55,280 --> 00:38:57,080 that I started off with. 777 00:38:57,080 --> 00:39:00,880 How do we combine data and storytelling? 778 00:39:00,880 --> 00:39:05,280 And one format that we use also in teaching is data stories. 779 00:39:05,280 --> 00:39:06,800 This is nothing that we invented. 780 00:39:06,800 --> 00:39:10,640 Data stories is a thing, just not in history. 781 00:39:10,640 --> 00:39:13,960 But in lots of different areas, there's 782 00:39:13,960 --> 00:39:18,360 data journalists that basically produce data stories every day. 783 00:39:18,360 --> 00:39:21,160 In businesses that sell data, they 784 00:39:21,160 --> 00:39:24,240 sell data by telling stories with that particular data 785 00:39:24,240 --> 00:39:27,160 to make a selling argument. 786 00:39:27,160 --> 00:39:30,000 So data stories is a format that is 787 00:39:30,000 --> 00:39:32,920 something that is very present ever since we 788 00:39:32,920 --> 00:39:37,240 have lots of data sets that need to be told or told about. 789 00:39:37,240 --> 00:39:40,640 And we tried to use data sets in a seminar 790 00:39:40,640 --> 00:39:45,880 again for a historical data story. 791 00:39:45,880 --> 00:39:50,400 And this data story is based on the Blumenbach online. 792 00:39:50,400 --> 00:39:57,400 Blumenbach is a-- let's call him ethnographer maybe, 793 00:39:57,400 --> 00:40:00,080 early or late 18th century. 794 00:40:00,080 --> 00:40:04,520 And he collected human remains, skulls in particular, 795 00:40:04,520 --> 00:40:11,640 in order to categorize humans in ABCD and so on categories. 796 00:40:11,640 --> 00:40:17,040 This is an event or a situation that is often also retold 797 00:40:17,040 --> 00:40:21,760 in stories of racism and the idea 798 00:40:21,760 --> 00:40:24,800 that you can see at people and put them 799 00:40:24,800 --> 00:40:25,800 into certain categories. 800 00:40:25,800 --> 00:40:28,680 And he collected these skulls all over the world. 801 00:40:28,680 --> 00:40:30,960 And the Blumenbach archive in Göttingen, 802 00:40:30,960 --> 00:40:33,100 they have his correspondence. 803 00:40:33,100 --> 00:40:35,520 They have the roots that-- 804 00:40:35,520 --> 00:40:38,200 they have the roots that skulls took. 805 00:40:38,200 --> 00:40:40,400 They have lots of metadata information on the skulls. 806 00:40:40,400 --> 00:40:43,000 And we used these data sets, or this particular data set, 807 00:40:43,000 --> 00:40:47,200 all surrounding the collection of Blumenbach 808 00:40:47,200 --> 00:40:53,520 to create a data story that basically looks like this. 809 00:40:53,520 --> 00:40:54,640 And this is German again. 810 00:40:54,640 --> 00:40:57,040 I apologize, but my students usually work in German. 811 00:40:57,040 --> 00:41:00,500 So we haven't yet had the chance to translate this. 812 00:41:00,500 --> 00:41:03,120 It's about collecting, describing, categorizing, 813 00:41:03,120 --> 00:41:06,020 about the-- 814 00:41:06,020 --> 00:41:07,920 I think I gave it an English title-- 815 00:41:07,920 --> 00:41:09,440 about the treatment of human remains 816 00:41:09,440 --> 00:41:11,140 at the beginning of modern science, 817 00:41:11,140 --> 00:41:13,800 because what the students were most interested in is, 818 00:41:13,800 --> 00:41:16,280 how does this collection of skulls 819 00:41:16,280 --> 00:41:18,200 relate to what we today think about how 820 00:41:18,200 --> 00:41:19,400 we deal with human remains? 821 00:41:19,400 --> 00:41:21,480 So this is the kind of story that they wanted to tell. 822 00:41:21,480 --> 00:41:22,600 And we had a data set. 823 00:41:22,600 --> 00:41:25,720 And then we tried to make a story out of it. 824 00:41:25,720 --> 00:41:28,060 And this is what the students did. 825 00:41:28,060 --> 00:41:32,500 So it has this opening teaser, basically. 826 00:41:32,500 --> 00:41:34,600 And then they decided that they wanted to track 827 00:41:34,600 --> 00:41:37,200 the movements of the skulls. 828 00:41:37,200 --> 00:41:39,920 And I think this is something that we in digital history, 829 00:41:39,920 --> 00:41:43,220 but maybe also, Manuel, in your group in computational history, 830 00:41:43,220 --> 00:41:47,600 what we often do is listen to someone who has an interest 831 00:41:47,600 --> 00:41:50,240 and has some material, and then try 832 00:41:50,240 --> 00:41:56,200 to design bits and pieces of computational methods 833 00:41:56,200 --> 00:42:00,600 or visual representations that can help in telling the story. 834 00:42:00,600 --> 00:42:03,260 Most of the time, the digital history project 835 00:42:03,260 --> 00:42:05,040 does not just consist of one method, 836 00:42:05,040 --> 00:42:07,500 but it's a set of methods that represent different steps 837 00:42:07,500 --> 00:42:09,460 in the research process. 838 00:42:09,460 --> 00:42:12,940 So what we did is we wanted to follow 839 00:42:12,940 --> 00:42:14,020 the journey of the skull. 840 00:42:14,020 --> 00:42:19,720 So we decided to do this with GIS methods, representations. 841 00:42:19,720 --> 00:42:24,060 So we designed interactive maps along three different points 842 00:42:24,060 --> 00:42:25,360 of the journeys. 843 00:42:25,360 --> 00:42:29,940 The first map is about where the skulls come from. 844 00:42:29,940 --> 00:42:32,500 And because I told you that, for me, it 845 00:42:32,500 --> 00:42:35,140 is important to make this interactive 846 00:42:35,140 --> 00:42:40,160 or to give the users possibilities to find 847 00:42:40,160 --> 00:42:42,720 their own interests. 848 00:42:42,720 --> 00:42:46,840 In the data sets, we have marked down all the locations 849 00:42:46,840 --> 00:42:49,840 where the skulls came from and linked them 850 00:42:49,840 --> 00:42:52,480 to the metadata information about the skull 851 00:42:52,480 --> 00:42:54,120 that we had in the database. 852 00:42:54,120 --> 00:42:56,160 So this is basically information retrieval, right? 853 00:42:56,160 --> 00:42:58,280 It's just pretty. 854 00:42:58,280 --> 00:43:00,920 You could do this also with a search field in the database, 855 00:43:00,920 --> 00:43:02,960 but this is more nice. 856 00:43:02,960 --> 00:43:06,000 It doesn't do much more than that. 857 00:43:06,000 --> 00:43:08,100 The second thing that students wanted to look at-- 858 00:43:08,100 --> 00:43:09,660 and in between these visualizations, 859 00:43:09,660 --> 00:43:11,200 they tell the story. 860 00:43:11,200 --> 00:43:13,260 They have the historical information, 861 00:43:13,260 --> 00:43:17,480 and they give the narrative, and they give you more information 862 00:43:17,480 --> 00:43:19,260 about certain skulls. 863 00:43:19,260 --> 00:43:22,640 And then they wanted to show the places 864 00:43:22,640 --> 00:43:27,280 where the skulls went to and the routes that they took. 865 00:43:27,280 --> 00:43:29,940 This is not well done because the routes all 866 00:43:29,940 --> 00:43:32,120 seem to be very straight. 867 00:43:32,120 --> 00:43:34,240 Obviously, that's not what happened. 868 00:43:34,240 --> 00:43:36,920 This is historically very incorrect, right? 869 00:43:36,920 --> 00:43:39,960 But this represents the points of data that we have. 870 00:43:39,960 --> 00:43:41,600 So the available data points often 871 00:43:41,600 --> 00:43:44,040 are where does the skull come from, 872 00:43:44,040 --> 00:43:46,800 and then we maybe have one or two brokers in between. 873 00:43:46,800 --> 00:43:49,040 So they were shipped to St. Petersburg, 874 00:43:49,040 --> 00:43:52,680 which is basically where, if you zoom into this, 875 00:43:52,680 --> 00:44:00,800 you can see this here, St. Petersburg, this person here. 876 00:44:00,800 --> 00:44:03,240 Just lots of connections. 877 00:44:03,240 --> 00:44:05,460 There's also lots of correspondence 878 00:44:05,460 --> 00:44:06,600 between him and Blumenbach. 879 00:44:06,600 --> 00:44:09,360 Niko Karpinski, he was one of the brokers 880 00:44:09,360 --> 00:44:15,140 that Blumenbach used to bring the skulls back to Göttingen. 881 00:44:15,140 --> 00:44:18,220 So we have some stations, but this is not 882 00:44:18,220 --> 00:44:21,600 a historically correct route that the skulls took, right? 883 00:44:21,600 --> 00:44:23,240 Part of the story that the students took 884 00:44:23,240 --> 00:44:26,800 was that one of the skulls, very old one, 885 00:44:26,800 --> 00:44:31,920 he was on the last bit of road just before Göttingen. 886 00:44:31,920 --> 00:44:33,600 The coach broke down. 887 00:44:33,600 --> 00:44:38,320 And then the box where the skull was in fell off the coach, 888 00:44:38,320 --> 00:44:39,820 and then it broke. 889 00:44:39,820 --> 00:44:41,960 So it had a very long story, but then it 890 00:44:41,960 --> 00:44:43,760 never arrived in Göttingen. 891 00:44:43,760 --> 00:44:47,920 This is what the students tell you in the story. 892 00:44:47,920 --> 00:44:50,920 And then the last step was that we were thinking about, OK, 893 00:44:50,920 --> 00:44:54,920 so we have the point where they start. 894 00:44:54,920 --> 00:44:57,720 We have the route that they take. 895 00:44:57,720 --> 00:45:00,280 Then they arrive in Göttingen, but then we would have just one 896 00:45:00,280 --> 00:45:02,320 bubble in Göttingen. 897 00:45:02,320 --> 00:45:04,480 So again, the logic of the visualization 898 00:45:04,480 --> 00:45:06,400 also helps you tell the story in a way 899 00:45:06,400 --> 00:45:08,960 that people can understand what you're trying to say. 900 00:45:08,960 --> 00:45:12,040 It's not an obvious sort of visual representation. 901 00:45:12,040 --> 00:45:15,200 It's always also a decision, part of the storytelling. 902 00:45:15,200 --> 00:45:17,800 So what we did was we went back to the information 903 00:45:17,800 --> 00:45:21,000 of where the skulls came from, but the colors and the letters 904 00:45:21,000 --> 00:45:23,880 represent the categories that Blumenbach later invented 905 00:45:23,880 --> 00:45:27,880 or that he built from the skulls. 906 00:45:27,880 --> 00:45:32,660 And then you can have your own thoughts about whether or not 907 00:45:32,660 --> 00:45:34,760 this works at all. 908 00:45:34,760 --> 00:45:36,360 But again, you can click on them, 909 00:45:36,360 --> 00:45:41,720 and then you get sort of the database information. 910 00:45:41,720 --> 00:45:45,600 So there is a way to still tell stories with data 911 00:45:45,600 --> 00:45:49,000 and integrate data in your stories 912 00:45:49,000 --> 00:45:50,780 with interactive visualizations. 913 00:45:50,780 --> 00:45:52,560 It's just a different format, right? 914 00:45:52,560 --> 00:45:54,920 It's not telling stories. 915 00:45:54,920 --> 00:45:56,420 It's not using data. 916 00:45:56,420 --> 00:46:00,760 It's just trying to explore ways to combine the two. 917 00:46:00,760 --> 00:46:01,880 Yeah, OK. 918 00:46:01,880 --> 00:46:04,080 I'm going to do this quickly because this is boring. 919 00:46:04,080 --> 00:46:05,200 This is what I do every day. 920 00:46:05,200 --> 00:46:07,240 So for me, it's boring, at least. 921 00:46:07,240 --> 00:46:11,000 I work on court records, medieval court records, 922 00:46:11,000 --> 00:46:13,560 from the so-called justices in Ayr. 923 00:46:13,560 --> 00:46:16,840 That's basically the king's judges in the 11th century. 924 00:46:16,840 --> 00:46:20,220 They started to not only stay at Westminster Court 925 00:46:20,220 --> 00:46:22,920 and receive all the litigants at Westminster, 926 00:46:22,920 --> 00:46:25,560 but they started to go to the people 927 00:46:25,560 --> 00:46:27,660 and bring them the king's justice. 928 00:46:27,660 --> 00:46:29,280 That's more or less the story. 929 00:46:29,280 --> 00:46:32,500 And they have this very long way that they covered. 930 00:46:32,500 --> 00:46:38,280 This took about two years to go from one end to the other. 931 00:46:38,280 --> 00:46:41,400 And this is-- I'm just going to show you this quickly. 932 00:46:41,400 --> 00:46:45,540 This is now an example for not using existing data, 933 00:46:45,540 --> 00:46:48,580 but me being a medieval historian interested in court 934 00:46:48,580 --> 00:46:52,180 records, trying to turn this into data that I can then 935 00:46:52,180 --> 00:46:55,480 use with quantitative methods. 936 00:46:55,480 --> 00:46:59,960 So what we do is we train machine learning algorithms 937 00:46:59,960 --> 00:47:03,240 in reading the scripts or transcribing 938 00:47:03,240 --> 00:47:05,720 the handwritten scripts. 939 00:47:05,720 --> 00:47:07,720 We turn this into machine readable text. 940 00:47:07,720 --> 00:47:10,080 And then we use lots of different natural language 941 00:47:10,080 --> 00:47:13,800 processing tools to treat the texts with. 942 00:47:13,800 --> 00:47:16,360 So named entity recognition, automatic recognition 943 00:47:16,360 --> 00:47:19,000 of person names, of locations, for example, 944 00:47:19,000 --> 00:47:21,760 trying to then build networks. 945 00:47:21,760 --> 00:47:24,360 We try to look at or to understand 946 00:47:24,360 --> 00:47:25,760 what are the legal terms that are 947 00:47:25,760 --> 00:47:29,160 present in the earlier records versus the later records. 948 00:47:29,160 --> 00:47:33,720 So trying to find how do these different legal categories 949 00:47:33,720 --> 00:47:35,380 stabilize over time. 950 00:47:35,380 --> 00:47:38,360 And because I have a big corpus, they 951 00:47:38,360 --> 00:47:42,600 started to collect these kinds of court rolls from about 1180 952 00:47:42,600 --> 00:47:44,340 until the 19th century. 953 00:47:44,340 --> 00:47:47,440 I'm going to stop somewhere way before that. 954 00:47:47,440 --> 00:47:51,040 But you could do this in a really large scale study 955 00:47:51,040 --> 00:47:53,760 if you had the tools, the time, the resources 956 00:47:53,760 --> 00:47:57,680 to digitize it all in a way that would make this possible. 957 00:47:57,680 --> 00:48:02,360 And we deal with metadata and textual data when we do this. 958 00:48:02,360 --> 00:48:04,120 We also deal with person names. 959 00:48:04,120 --> 00:48:09,760 And what you usually do if you identify people, 960 00:48:09,760 --> 00:48:10,940 you build networks, right? 961 00:48:10,940 --> 00:48:13,600 Because this is a go-to method if you 962 00:48:13,600 --> 00:48:16,240 want to think about the relation again. 963 00:48:16,240 --> 00:48:18,980 I think I started off with-- 964 00:48:18,980 --> 00:48:20,740 also, I made this point earlier that I'm 965 00:48:20,740 --> 00:48:22,460 about relations and relational history. 966 00:48:22,460 --> 00:48:25,300 So I did this little plot. 967 00:48:25,300 --> 00:48:26,360 It's not one network. 968 00:48:26,360 --> 00:48:28,820 As you can see, it's lots of really tiny networks 969 00:48:28,820 --> 00:48:32,880 that just connect three or two or four nodes with each other. 970 00:48:32,880 --> 00:48:37,720 What you can see here, every gray node is a court case. 971 00:48:37,720 --> 00:48:39,220 And the other nodes are the people 972 00:48:39,220 --> 00:48:40,380 involved in the court case. 973 00:48:40,380 --> 00:48:42,720 So it's usually two people, one case. 974 00:48:42,720 --> 00:48:45,960 Sometimes it's more, but that's sort of the usual thing. 975 00:48:45,960 --> 00:48:48,900 And then the color of the nodes tell you 976 00:48:48,900 --> 00:48:52,480 which social strata or category these people come from. 977 00:48:52,480 --> 00:48:56,220 Because I'm interested to see how the social strata that I 978 00:48:56,220 --> 00:48:58,240 am expecting in a medieval society 979 00:48:58,240 --> 00:49:00,680 is represented in these court records. 980 00:49:00,680 --> 00:49:02,500 Is it villages against villagers? 981 00:49:02,500 --> 00:49:04,500 Is it villages against regional landholders? 982 00:49:04,500 --> 00:49:06,900 Is it knights against clerics? 983 00:49:06,900 --> 00:49:11,800 What are the social connections that 984 00:49:11,800 --> 00:49:13,500 are built in the situation in court? 985 00:49:13,500 --> 00:49:15,780 Because these are people that usually maybe not talk 986 00:49:15,780 --> 00:49:17,340 to each other on the street. 987 00:49:17,340 --> 00:49:19,580 But if they have a legal battle, they need to. 988 00:49:19,580 --> 00:49:25,580 So what is the area the court, as an area 989 00:49:25,580 --> 00:49:27,620 of social interaction does to this? 990 00:49:27,620 --> 00:49:33,140 And as you can see, these cases connect different people. 991 00:49:33,140 --> 00:49:37,340 But sometimes, people also connect cases. 992 00:49:37,340 --> 00:49:41,580 And sometimes, only people connect to each other. 993 00:49:41,580 --> 00:49:43,440 And these are different observations 994 00:49:43,440 --> 00:49:46,860 that I can make on a pattern level, pattern recognition 995 00:49:46,860 --> 00:49:49,100 mechanisms that I can use that help 996 00:49:49,100 --> 00:49:52,540 me ask questions that will then bring me back to the text. 997 00:49:52,540 --> 00:49:54,980 So this is not a result of any kind. 998 00:49:54,980 --> 00:49:57,920 This is just one further step in the research process. 999 00:49:57,920 --> 00:49:59,740 If I'm interested in the kinds of networks 1000 00:49:59,740 --> 00:50:03,560 that are represented, I would now go back and see, OK, 1001 00:50:03,560 --> 00:50:05,100 this one here, is this a mistake? 1002 00:50:05,100 --> 00:50:08,940 Is there just no case mentioned? 1003 00:50:08,940 --> 00:50:10,400 I did go back. 1004 00:50:10,400 --> 00:50:13,540 And it's actually a retrial of a previous case. 1005 00:50:13,540 --> 00:50:16,180 So the case is not mentioned anymore. 1006 00:50:16,180 --> 00:50:17,280 It's just about the people. 1007 00:50:17,280 --> 00:50:20,780 They come back to court like a week later and just continue. 1008 00:50:20,780 --> 00:50:24,380 But because it's further down in the document, 1009 00:50:24,380 --> 00:50:27,380 it appears as a single entry. 1010 00:50:27,380 --> 00:50:30,460 So I learned something about the structure of the document 1011 00:50:30,460 --> 00:50:34,300 also from the personal network. 1012 00:50:34,300 --> 00:50:37,060 And I think this is very important that you learn 1013 00:50:37,060 --> 00:50:38,900 how to read the visualizations and learn 1014 00:50:38,900 --> 00:50:40,380 how to read the kinds of data points 1015 00:50:40,380 --> 00:50:44,200 that you have in order to get all the factors into your story 1016 00:50:44,200 --> 00:50:45,280 again. 1017 00:50:45,280 --> 00:50:47,200 Because I think that-- or I observe 1018 00:50:47,200 --> 00:50:49,840 that one of the main arguments against working with data 1019 00:50:49,840 --> 00:50:53,080 often is, what do you do with context? 1020 00:50:53,080 --> 00:50:56,420 If you isolate something as a data point, 1021 00:50:56,420 --> 00:50:58,440 how do you get the context information back in, 1022 00:50:58,440 --> 00:51:00,220 which is one of the most important things 1023 00:51:00,220 --> 00:51:02,240 for historical interpretations? 1024 00:51:02,240 --> 00:51:05,900 And this is one way that I use these visualizations that 1025 00:51:05,900 --> 00:51:08,640 bring me back to the context and that I understand the way 1026 00:51:08,640 --> 00:51:10,680 that the documents are structured, for example. 1027 00:51:10,680 --> 00:51:18,920 So for me, digital history and also the way 1028 00:51:18,920 --> 00:51:24,200 that data is used in historical storytelling 1029 00:51:24,200 --> 00:51:28,120 is quantitative methods. 1030 00:51:28,120 --> 00:51:30,440 I would rather call myself a data historian 1031 00:51:30,440 --> 00:51:31,520 than a digital historian. 1032 00:51:31,520 --> 00:51:34,120 Because I think the differentiation is much clearer 1033 00:51:34,120 --> 00:51:36,540 if I do that than just say I work digitally, 1034 00:51:36,540 --> 00:51:40,000 because who doesn't? 1035 00:51:40,000 --> 00:51:42,600 But just a few points on this slide. 1036 00:51:42,600 --> 00:51:47,080 So when I use these kinds of methods for interpretations 1037 00:51:47,080 --> 00:51:51,600 of data that I created, it is important to think about this 1038 00:51:51,600 --> 00:51:54,720 as methods of pattern recognition. 1039 00:51:54,720 --> 00:51:57,160 They allow you to make observations that you cannot 1040 00:51:57,160 --> 00:51:59,280 make when you just read a text. 1041 00:51:59,280 --> 00:52:01,640 Because texts are organized sequentially. 1042 00:52:01,640 --> 00:52:02,880 It's linear. 1043 00:52:02,880 --> 00:52:04,020 It has a certain order. 1044 00:52:04,020 --> 00:52:06,000 It represents-- and it's not-- 1045 00:52:06,000 --> 00:52:07,980 I'm not saying that you don't read anymore. 1046 00:52:07,980 --> 00:52:10,200 It's just a different perspective 1047 00:52:10,200 --> 00:52:13,860 that you then have to bring together again. 1048 00:52:13,860 --> 00:52:15,900 And it's about modeling of research questions 1049 00:52:15,900 --> 00:52:18,060 and thinking about how can I operationalize 1050 00:52:18,060 --> 00:52:23,020 certain sets of my-- or certain questions within my research 1051 00:52:23,020 --> 00:52:26,920 focus in a way that I can use formalized and quantitative 1052 00:52:26,920 --> 00:52:28,580 methods. 1053 00:52:28,580 --> 00:52:31,440 These networks don't make sense for all the questions 1054 00:52:31,440 --> 00:52:34,420 that you could ask for the court records, right? 1055 00:52:34,420 --> 00:52:38,640 It's just one particular set. 1056 00:52:38,640 --> 00:52:41,480 Unfortunately, this means that research questions 1057 00:52:41,480 --> 00:52:44,860 need to be formalized in order to be able to do this. 1058 00:52:44,860 --> 00:52:46,580 And this is true for any kind of data set. 1059 00:52:46,580 --> 00:52:49,360 If you want to match a data set, even social strata data 1060 00:52:49,360 --> 00:52:53,360 or social surveys that are already existing, 1061 00:52:53,360 --> 00:52:56,200 you kind of have to match your research interest 1062 00:52:56,200 --> 00:52:59,140 with something that is very formalized, that 1063 00:52:59,140 --> 00:53:01,400 is very strictly organized. 1064 00:53:01,400 --> 00:53:02,900 And sometimes our questions are not. 1065 00:53:02,900 --> 00:53:06,700 Maybe, again, there's a chance to bring these two kinds 1066 00:53:06,700 --> 00:53:09,540 of perspectives together. 1067 00:53:09,540 --> 00:53:12,620 And I say it again because you always 1068 00:53:12,620 --> 00:53:15,100 have to make people aware. 1069 00:53:15,100 --> 00:53:18,380 Formalization needs documentation. 1070 00:53:18,380 --> 00:53:21,220 You have to really make sure that everybody understands 1071 00:53:21,220 --> 00:53:23,220 the decisions that you made so that they can use 1072 00:53:23,220 --> 00:53:25,520 the data that you produce. 1073 00:53:25,520 --> 00:53:26,900 It's an iterative process. 1074 00:53:26,900 --> 00:53:29,340 It's just part of the research process. 1075 00:53:29,340 --> 00:53:31,760 Visualizations and these kinds of things that I'm showing you 1076 00:53:31,760 --> 00:53:35,840 is never, for me, the end of the line. 1077 00:53:35,840 --> 00:53:38,120 I produce a new set of information 1078 00:53:38,120 --> 00:53:42,540 that I then use in order to combine quantitative modeling 1079 00:53:42,540 --> 00:53:44,240 and qualitative interpretation. 1080 00:53:44,240 --> 00:53:52,680 This is the virtual reality application now. 1081 00:53:52,680 --> 00:53:57,040 Maybe we can do this over drinks. 1082 00:53:57,040 --> 00:54:00,420 I thought I would show you this. 1083 00:54:00,420 --> 00:54:02,000 I will show you later. 1084 00:54:02,000 --> 00:54:03,840 And I will solve the riddle of the acronym. 1085 00:54:03,840 --> 00:54:13,280 What we try to do is in order to link quantitative observations 1086 00:54:13,280 --> 00:54:16,880 and sort of data visualization and information 1087 00:54:16,880 --> 00:54:23,320 I get from looking at things to understanding that I am always 1088 00:54:23,320 --> 00:54:27,420 situated is basically exactly this. 1089 00:54:27,420 --> 00:54:30,760 So you are in a space that looks like space. 1090 00:54:30,760 --> 00:54:36,560 And you can move around these knowledge graphs. 1091 00:54:36,560 --> 00:54:39,800 This is taken-- so we started to model this 1092 00:54:39,800 --> 00:54:42,640 according to the [SPEAKING GERMAN] 1093 00:54:42,640 --> 00:54:44,840 that I showed you, or [SPEAKING GERMAN] 1094 00:54:44,840 --> 00:54:46,480 The problem is that we understood 1095 00:54:46,480 --> 00:54:49,920 that the way that these networks are built are totally random. 1096 00:54:49,920 --> 00:54:52,120 Because there's no-- 1097 00:54:52,120 --> 00:54:58,180 I mean, it's sort of computational, but they calculated 1098 00:54:58,180 --> 00:55:03,260 the size of the nodes, which represented 1099 00:55:03,260 --> 00:55:08,260 the semantic closeness of two events to each other. 1100 00:55:08,260 --> 00:55:12,620 And they didn't use the same keyword every time. 1101 00:55:12,620 --> 00:55:15,580 They had a very complex set of calculations, 1102 00:55:15,580 --> 00:55:17,540 what is an important keyword, and it 1103 00:55:17,540 --> 00:55:20,940 was just not understandable, which I translate as random. 1104 00:55:20,940 --> 00:55:23,820 It's obviously not random at all, 1105 00:55:23,820 --> 00:55:25,840 but it doesn't have a good documentation, 1106 00:55:25,840 --> 00:55:29,600 so you can't really understand what the relations are. 1107 00:55:29,600 --> 00:55:34,080 Wikipedia, Wikidata, and DBpedia have a very much better 1108 00:55:34,080 --> 00:55:36,400 documentation of the kinds of relations that they make. 1109 00:55:36,400 --> 00:55:38,000 So this is basically what you see here. 1110 00:55:38,000 --> 00:55:42,760 This is a medieval 100 years war knowledge graph 1111 00:55:42,760 --> 00:55:49,300 taken from the DBpedia database. 1112 00:55:49,300 --> 00:55:51,380 And if we have some time later in the discussion, 1113 00:55:51,380 --> 00:55:52,740 maybe I can also show you this. 1114 00:55:52,740 --> 00:55:56,820 But to solve the riddle of the etard is, I just point to this. 1115 00:55:56,820 --> 00:56:05,380 And actually, to give you a little anecdote, 1116 00:56:05,380 --> 00:56:08,980 so the professor that I started this project with, 1117 00:56:08,980 --> 00:56:11,500 he said he was from computer vision and computer 1118 00:56:11,500 --> 00:56:12,100 visualization. 1119 00:56:12,100 --> 00:56:15,340 He did the whole 3D calculations. 1120 00:56:15,340 --> 00:56:17,820 And he said, OK, Silke, we can do this, 1121 00:56:17,820 --> 00:56:20,900 but it has to be Dr. Who. 1122 00:56:20,900 --> 00:56:23,780 So we knew the thing had to be called TARDIS 1123 00:56:23,780 --> 00:56:25,780 in some kind of combination. 1124 00:56:25,780 --> 00:56:26,940 And it is now-- 1125 00:56:26,940 --> 00:56:29,500 it was the Exploración-- 1126 00:56:29,500 --> 00:56:33,260 yes. 1127 00:56:33,260 --> 00:56:38,060 Sorry, I'm laughing about myself. 1128 00:56:38,060 --> 00:56:44,420 It's the Exploración Temporale und Reumlich Haddad 1129 00:56:44,420 --> 00:56:45,500 in Emeziven Cenarion. 1130 00:56:45,500 --> 00:56:47,500 So it actually has a title that later 1131 00:56:47,500 --> 00:56:50,220 then came to this acronym, but the acronym was first. 1132 00:56:50,220 --> 00:56:53,220 And I do believe that lots of projects work like that. 1133 00:56:53,220 --> 00:56:57,300 You have an acronym, and you make the title work to them. 1134 00:56:57,300 --> 00:56:59,940 So maybe we can have a look at this later. 1135 00:56:59,940 --> 00:57:01,340 To come to an end, because I think 1136 00:57:01,340 --> 00:57:06,580 I've been kind of approaching the end of this, 1137 00:57:06,580 --> 00:57:08,860 do numbers tell stories? 1138 00:57:08,860 --> 00:57:12,060 I don't know whether this is a yes or no answer, 1139 00:57:12,060 --> 00:57:15,060 as most historical questions are not a yes or no answer. 1140 00:57:15,060 --> 00:57:17,220 I can maybe leave it like this. 1141 00:57:17,220 --> 00:57:19,620 What I wanted to point out, and what is important to me, 1142 00:57:19,620 --> 00:57:22,600 is that stories and narrations have many forms. 1143 00:57:22,600 --> 00:57:25,780 I showed you a format of data stories, for example. 1144 00:57:25,780 --> 00:57:29,220 But a data set also is basically a representation of a story. 1145 00:57:29,220 --> 00:57:32,260 And so far, as it is linked to certain presumptions 1146 00:57:32,260 --> 00:57:33,740 representing certain perspectives, 1147 00:57:33,740 --> 00:57:39,300 it's just not neutral, because it's in the world, right? 1148 00:57:39,300 --> 00:57:41,900 Important for me also, because I think a lot 1149 00:57:41,900 --> 00:57:44,180 and write about the role that visualizations 1150 00:57:44,180 --> 00:57:48,300 have in this whole process of interpreting 1151 00:57:48,300 --> 00:57:51,500 data sets for historians. 1152 00:57:51,500 --> 00:57:53,140 Visualizations also tell stories. 1153 00:57:53,140 --> 00:57:57,480 You always decide what you can see and what you need to hide. 1154 00:57:57,480 --> 00:57:59,640 They represent data models and make certain information 1155 00:57:59,640 --> 00:58:01,820 visible and others invisible. 1156 00:58:01,820 --> 00:58:03,540 And this is what narratives do, right? 1157 00:58:03,540 --> 00:58:04,980 So you tell a story. 1158 00:58:04,980 --> 00:58:07,060 You pick out something that is really important, 1159 00:58:07,060 --> 00:58:09,740 maybe because it's exemplary or representative, 1160 00:58:09,740 --> 00:58:11,940 or maybe it's the outlier, and that's 1161 00:58:11,940 --> 00:58:13,500 what's interesting for you. 1162 00:58:13,500 --> 00:58:15,700 But you will never give us the whole picture. 1163 00:58:15,700 --> 00:58:17,300 And this is what visualizations do. 1164 00:58:17,300 --> 00:58:18,460 They give you patterns. 1165 00:58:18,460 --> 00:58:20,660 And then you can zoom in. 1166 00:58:20,660 --> 00:58:22,900 You can zoom out. 1167 00:58:22,900 --> 00:58:25,860 You can include something, exclude something. 1168 00:58:25,860 --> 00:58:29,100 You can say, just give me the unknown data points, 1169 00:58:29,100 --> 00:58:32,260 because I'm interested in all the facts that we don't have, 1170 00:58:32,260 --> 00:58:34,820 and try to think about why we don't have this. 1171 00:58:34,820 --> 00:58:38,040 So I think this also helps us to give a new perspective 1172 00:58:38,040 --> 00:58:41,100 to our own stories. 1173 00:58:41,100 --> 00:58:42,980 Visualizations help with the analysis of data, 1174 00:58:42,980 --> 00:58:46,660 because it just helps you to see. 1175 00:58:46,660 --> 00:58:50,860 But it's also really important to think about visual literacy 1176 00:58:50,860 --> 00:58:54,660 to understand what happens when you see, right? 1177 00:58:54,660 --> 00:58:57,740 And to make sure, again, that people 1178 00:58:57,740 --> 00:59:03,180 know that this is a game of hide and seek, basically. 1179 00:59:03,180 --> 00:59:05,420 So figuring out the past-- 1180 00:59:05,420 --> 00:59:08,400 and with this, I don't mean the book, 1181 00:59:08,400 --> 00:59:10,260 because I think I said lots about the book. 1182 00:59:10,260 --> 00:59:13,540 It's really interesting. 1183 00:59:13,540 --> 00:59:18,880 But figuring out the past maybe as a metaphorical phrase 1184 00:59:18,880 --> 00:59:20,380 is a multilayered process involving 1185 00:59:20,380 --> 00:59:21,580 different steps and methods. 1186 00:59:21,580 --> 00:59:23,580 And some of them can be computational. 1187 00:59:23,580 --> 00:59:26,180 Some of them can relate to data. 1188 00:59:26,180 --> 00:59:28,540 And you use the data, the stories 1189 00:59:28,540 --> 00:59:32,940 that are sort of encompassed in data sets 1190 00:59:32,940 --> 00:59:35,620 to give you a new insight, maybe. 1191 00:59:35,620 --> 00:59:37,740 Numbers can be part of the stories we tell, 1192 00:59:37,740 --> 00:59:39,760 and they can help with plausibleization. 1193 00:59:39,760 --> 00:59:45,500 I think in the little text that I sent you to announce the talk, 1194 00:59:45,500 --> 00:59:49,020 I said that this is something that we sort of expect 1195 00:59:49,020 --> 00:59:50,580 from data sets nowadays. 1196 00:59:50,580 --> 00:59:54,260 If you can give it-- if you can assign a number to something, 1197 00:59:54,260 --> 00:59:57,300 say 80% of people say this and that, 1198 00:59:57,300 --> 01:00:02,080 this is how people plausibleize the story, right? 1199 01:00:02,080 --> 01:00:04,300 It's more plausible if more people say this. 1200 01:00:04,300 --> 01:00:07,860 Or a YouTube video is especially true, 1201 01:00:07,860 --> 01:00:09,960 like if you look at the history channels, when 1202 01:00:09,960 --> 01:00:12,140 lots of people like it. 1203 01:00:12,140 --> 01:00:13,140 I don't know. 1204 01:00:13,140 --> 01:00:15,780 It's a weird way to validate this, 1205 01:00:15,780 --> 01:00:18,160 but this is what happens, right? 1206 01:00:18,160 --> 01:00:20,700 If you Google something, just because something 1207 01:00:20,700 --> 01:00:24,600 is sort of the top answer doesn't make it true or false 1208 01:00:24,600 --> 01:00:25,740 or anything. 1209 01:00:25,740 --> 01:00:28,580 So these mechanisms are also numbers 1210 01:00:28,580 --> 01:00:32,900 that sort of influence the way that we understand the world, 1211 01:00:32,900 --> 01:00:36,180 the way that people treat facts, the way that you 1212 01:00:36,180 --> 01:00:39,940 build plausible narratives without ever documenting 1213 01:00:39,940 --> 01:00:45,060 the way that led you to that particular plausible story. 1214 01:00:45,060 --> 01:00:47,540 And authorship and perspective are inscribed 1215 01:00:47,540 --> 01:00:48,980 in all forms of storytelling. 1216 01:00:48,980 --> 01:00:50,500 And this is the point that I made 1217 01:00:50,500 --> 01:00:53,300 with the interactive design of visualizations 1218 01:00:53,300 --> 01:00:55,540 and also of our virtual reality application. 1219 01:00:55,540 --> 01:00:57,020 I think it's important to let people 1220 01:00:57,020 --> 01:01:00,640 know that they can direct their own perspective, 1221 01:01:00,640 --> 01:01:05,140 that they can find their own view on things, 1222 01:01:05,140 --> 01:01:08,460 and that it's necessary that if they can do this, 1223 01:01:08,460 --> 01:01:10,700 then probably the people designing the data set 1224 01:01:10,700 --> 01:01:12,660 have already done this also. 1225 01:01:12,660 --> 01:01:16,820 So it's not just a neutral relationship. 1226 01:01:16,820 --> 01:01:20,420 So yes, numbers tell stories sometimes. 1227 01:01:20,420 --> 01:01:23,900 But sometimes also stories produce numbers. 1228 01:01:23,900 --> 01:01:26,580 And I hope that you enjoyed this little ride through what 1229 01:01:26,580 --> 01:01:29,340 I think data and stories and digital history 1230 01:01:29,340 --> 01:01:30,700 have to tell each other. 1231 01:01:30,700 --> 01:01:31,500 Thank you very much. 1232 01:01:31,500 --> 01:01:34,940 [APPLAUSE] 1233 01:01:34,940 --> 01:01:41,920 All right, I think it's now time for discussion. 1234 01:01:41,920 --> 01:01:46,820 And I forgot an important hint in the introductory part. 1235 01:01:46,820 --> 01:01:49,340 Please don't run away after the discussion 1236 01:01:49,340 --> 01:01:52,920 because there will be food and drinks afterwards nearby, 1237 01:01:52,920 --> 01:01:54,260 I guess. 1238 01:01:54,260 --> 01:01:57,780 So yeah, and we can also continue the discussion, 1239 01:01:57,780 --> 01:02:00,220 as you mentioned, with foods and drinks. 1240 01:02:00,220 --> 01:02:02,020 But first, discussion. 1241 01:02:02,020 --> 01:02:06,580 Thank you very much for this very interesting talk. 1242 01:02:06,580 --> 01:02:10,260 I was wondering also, as a historian 1243 01:02:10,260 --> 01:02:12,340 and as a practitioner of digital history, 1244 01:02:12,340 --> 01:02:20,820 if the problem of ambiguity isn't, in part, 1245 01:02:20,820 --> 01:02:25,960 one that is self-created by speaking of data. 1246 01:02:25,960 --> 01:02:28,700 Because if you go back to history, 1247 01:02:28,700 --> 01:02:34,340 and one will talk about visualizing sources 1248 01:02:34,340 --> 01:02:37,740 which come from different standpoints, 1249 01:02:37,740 --> 01:02:39,820 different institutions, and so on, 1250 01:02:39,820 --> 01:02:41,580 that would always be in our mind. 1251 01:02:41,580 --> 01:02:50,300 If one talks about data, it has this unambiguous, almost 1252 01:02:50,300 --> 01:02:55,020 positive, positivist connotation, 1253 01:02:55,020 --> 01:02:57,780 which one then needs to question in a second step? 1254 01:02:57,780 --> 01:03:04,360 But why do you speak about data and not 1255 01:03:04,360 --> 01:03:09,820 sources and this connection? 1256 01:03:09,820 --> 01:03:10,380 Good point. 1257 01:03:10,380 --> 01:03:13,620 Again, it's all about perspective. 1258 01:03:13,620 --> 01:03:19,140 So I stress data because I think that the way that I work 1259 01:03:19,140 --> 01:03:22,060 with the historical material that I call my sources 1260 01:03:22,060 --> 01:03:25,100 is different to what other people do when 1261 01:03:25,100 --> 01:03:31,100 they say I read a source because I change the object. 1262 01:03:31,100 --> 01:03:34,740 If you think about the medieval manuscript that I showed you, 1263 01:03:34,740 --> 01:03:37,220 the core protocol, taking a picture 1264 01:03:37,220 --> 01:03:41,340 is already a representation and a move into another medium 1265 01:03:41,340 --> 01:03:42,740 because I don't have the parchment. 1266 01:03:42,740 --> 01:03:46,900 I don't see the nicks and nacks and all the things 1267 01:03:46,900 --> 01:03:49,460 on the margins and so on. 1268 01:03:49,460 --> 01:03:52,860 But if you turn this into a TXT file 1269 01:03:52,860 --> 01:03:55,340 and then use natural language processing methods where 1270 01:03:55,340 --> 01:03:57,960 you don't really see a sequential text anymore, 1271 01:03:57,960 --> 01:04:00,340 but you have this bag of words approach 1272 01:04:00,340 --> 01:04:03,100 where all links and contexts are deleted 1273 01:04:03,100 --> 01:04:05,700 and you use tokens as data points, 1274 01:04:05,700 --> 01:04:09,180 I think that your research object has really changed. 1275 01:04:09,180 --> 01:04:13,260 It's not a charter or a court record anymore. 1276 01:04:13,260 --> 01:04:16,340 It was derived from that originally. 1277 01:04:16,340 --> 01:04:20,500 So that's why I always say it is important that we clarify 1278 01:04:20,500 --> 01:04:24,220 for us as historians that there's a difference 1279 01:04:24,220 --> 01:04:25,620 between source and data. 1280 01:04:25,620 --> 01:04:27,740 And I think this is also a discussion that I learned 1281 01:04:27,740 --> 01:04:30,700 from research data management, that it's really 1282 01:04:30,700 --> 01:04:33,160 important to think about there's a broad understanding 1283 01:04:33,160 --> 01:04:35,060 of research data where you say the source is 1284 01:04:35,060 --> 01:04:36,140 part of the research data. 1285 01:04:36,140 --> 01:04:37,600 And there's a narrow one where you 1286 01:04:37,600 --> 01:04:43,260 say it becomes research data once I do something with it. 1287 01:04:43,260 --> 01:04:46,100 So those would be the two main reasons. 1288 01:04:46,100 --> 01:04:48,660 I think it's a different object. 1289 01:04:48,660 --> 01:04:52,860 And it is a different object because I changed it. 1290 01:04:52,860 --> 01:04:56,300 So maybe capta would be a better phrase in general, 1291 01:04:56,300 --> 01:04:58,420 but that's why I would always say this is data 1292 01:04:58,420 --> 01:05:00,460 and it's not just a source. 1293 01:05:00,460 --> 01:05:02,620 But you could go back, or maybe you 1294 01:05:02,620 --> 01:05:06,260 need to go back and say it's not the original source, 1295 01:05:06,260 --> 01:05:10,220 but I can treat it as if it were some kind of source 1296 01:05:10,220 --> 01:05:11,660 because I have to be critical. 1297 01:05:11,660 --> 01:05:14,500 So I have to know who did it, what is the author, 1298 01:05:14,500 --> 01:05:16,020 all the questions that we usually ask 1299 01:05:16,020 --> 01:05:22,420 when we use the material from archives or archival research. 1300 01:05:22,420 --> 01:05:23,420 More questions? 1301 01:05:23,420 --> 01:05:32,780 I might take the chance to ask a question myself. 1302 01:05:32,780 --> 01:05:33,820 So great talk. 1303 01:05:33,820 --> 01:05:34,700 Thank you. 1304 01:05:34,700 --> 01:05:36,860 I was really intrigued by the-- 1305 01:05:36,860 --> 01:05:38,820 I mean, it's kind of trivial, I guess. 1306 01:05:38,820 --> 01:05:42,340 But I mean, at one point, you said-- 1307 01:05:42,340 --> 01:05:45,260 so it was about using data to tell stories, right? 1308 01:05:45,260 --> 01:05:47,740 And when historians use data, they 1309 01:05:47,740 --> 01:05:49,620 tell a story from a certain perspective. 1310 01:05:49,620 --> 01:05:53,500 And then you switch back to the interactive visualizations 1311 01:05:53,500 --> 01:05:55,420 and made the point that these visualizations 1312 01:05:55,420 --> 01:05:59,180 can change the perspective in an interactive way. 1313 01:05:59,180 --> 01:06:01,700 And I mean, it's not just in digital humanities 1314 01:06:01,700 --> 01:06:02,820 interactive visualizations. 1315 01:06:02,820 --> 01:06:04,460 It's basically the whole toolbox we 1316 01:06:04,460 --> 01:06:06,060 have in digital humanities, right? 1317 01:06:06,060 --> 01:06:08,420 We have many ways to change parameters 1318 01:06:08,420 --> 01:06:11,340 in topic models or styleometry. 1319 01:06:11,340 --> 01:06:12,980 And I mean, in fact, I think we've 1320 01:06:12,980 --> 01:06:14,980 seen a lot of abuse of these methods 1321 01:06:14,980 --> 01:06:19,820 because very often people change these parameters in a black box 1322 01:06:19,820 --> 01:06:22,780 until they get a perspective they like. 1323 01:06:22,780 --> 01:06:26,980 And it's very often very random because they 1324 01:06:26,980 --> 01:06:33,660 change the perspective, and they manipulate parameters. 1325 01:06:33,660 --> 01:06:37,280 But they don't really understand what they changed 1326 01:06:37,280 --> 01:06:40,620 and how a specific distance measure influences 1327 01:06:40,620 --> 01:06:41,700 the visualization. 1328 01:06:41,700 --> 01:06:43,340 And I was just wondering-- 1329 01:06:43,340 --> 01:06:48,020 so you mentioned how important in history studies 1330 01:06:48,020 --> 01:06:49,840 source criticism is. 1331 01:06:49,840 --> 01:06:54,020 And I just wonder what you think about the importance of also 1332 01:06:54,020 --> 01:06:56,140 being critical about tools and knowing 1333 01:06:56,140 --> 01:07:00,660 how switching certain parameters and changing certain, well, 1334 01:07:00,660 --> 01:07:04,740 variables influences the actual perspective on the data. 1335 01:07:04,740 --> 01:07:05,460 Yeah. 1336 01:07:05,460 --> 01:07:07,420 I think it's really eminently critical 1337 01:07:07,420 --> 01:07:12,020 that you also learn about tool criticism and think about-- 1338 01:07:12,020 --> 01:07:16,220 because again-- and this is so weird. 1339 01:07:16,220 --> 01:07:18,780 We've been having a conversation earlier today 1340 01:07:18,780 --> 01:07:20,700 that we read articles from 10 years ago. 1341 01:07:20,700 --> 01:07:23,820 And we have the impression that we talk about the same problems 1342 01:07:23,820 --> 01:07:26,940 even though we have all this time past. 1343 01:07:26,940 --> 01:07:32,540 And I think that-- 1344 01:07:32,540 --> 01:07:34,740 I have lots of conversations with students 1345 01:07:34,740 --> 01:07:36,540 but also with other researchers about-- 1346 01:07:36,540 --> 01:07:40,220 and they have this expectation that data sets-- 1347 01:07:40,220 --> 01:07:42,540 anything that has to do with a computer 1348 01:07:42,540 --> 01:07:46,180 is more neutral than if they would read it and write 1349 01:07:46,180 --> 01:07:48,900 it themselves or if they could see the person that 1350 01:07:48,900 --> 01:07:51,620 is behind the tool, which I think 1351 01:07:51,620 --> 01:07:54,420 is one layer why tool criticism is so important. 1352 01:07:54,420 --> 01:07:56,260 Because of course, even creating the tool 1353 01:07:56,260 --> 01:07:59,540 and manipulating the tool in the way that you just said 1354 01:07:59,540 --> 01:08:01,940 is always intentional. 1355 01:08:01,940 --> 01:08:03,380 And it has a certain perspective. 1356 01:08:03,380 --> 01:08:04,980 So this is really, really important, 1357 01:08:04,980 --> 01:08:08,940 which is why in teaching and also in other contexts, 1358 01:08:08,940 --> 01:08:11,760 I really like to start exploring with something 1359 01:08:11,760 --> 01:08:14,820 like buoyant tools, which is a web application 1360 01:08:14,820 --> 01:08:16,620 that everybody can use. 1361 01:08:16,620 --> 01:08:18,340 It's an out-of-the-box tool that gives you 1362 01:08:18,340 --> 01:08:20,740 some understanding of how these quantitative methods 1363 01:08:20,740 --> 01:08:23,260 and visualizations work. 1364 01:08:23,260 --> 01:08:25,580 But it doesn't give you enough-- 1365 01:08:25,580 --> 01:08:27,660 like if you're interested, if you're hooked, 1366 01:08:27,660 --> 01:08:30,820 when you all see these beautiful visualizations, 1367 01:08:30,820 --> 01:08:33,700 doesn't give you the chance to then go back into and do 1368 01:08:33,700 --> 01:08:36,060 all the proper manipulation that you would actually 1369 01:08:36,060 --> 01:08:41,600 need to get what you're interested in. 1370 01:08:41,600 --> 01:08:43,900 But I think the process and understanding 1371 01:08:43,900 --> 01:08:48,620 that these tools just give you something that is a very 1372 01:08:48,620 --> 01:08:50,700 particular something and that you 1373 01:08:50,700 --> 01:08:53,020 have to understand how this comes into being 1374 01:08:53,020 --> 01:08:56,380 and then go back and learn about, OK, how does this work? 1375 01:08:56,380 --> 01:08:59,780 And again, for example, in research papers-- 1376 01:08:59,780 --> 01:09:03,860 and I think we should talk a lot about also data publications, 1377 01:09:03,860 --> 01:09:06,420 not just sort of the usual research paper 1378 01:09:06,420 --> 01:09:09,420 that historians write, but make it more about the data set, 1379 01:09:09,420 --> 01:09:13,100 make it more about the tools, and be more transparent 1380 01:09:13,100 --> 01:09:18,980 about the process and document, the several steps that we did. 1381 01:09:18,980 --> 01:09:22,260 Like, for example, my little collection of network graphs. 1382 01:09:22,260 --> 01:09:25,860 I could have just hidden the one that didn't have a case, 1383 01:09:25,860 --> 01:09:28,220 but then I wouldn't have seen something 1384 01:09:28,220 --> 01:09:31,300 that really helped me understand the material that these network 1385 01:09:31,300 --> 01:09:34,340 graphs were coming from. 1386 01:09:34,340 --> 01:09:38,940 So again, we have to have an understanding of the tools, 1387 01:09:38,940 --> 01:09:41,140 and we have to be critical about the way 1388 01:09:41,140 --> 01:09:43,540 that they were designed because some or most of the methods 1389 01:09:43,540 --> 01:09:45,540 were-- 1390 01:09:45,540 --> 01:09:47,500 the trick is none of them-- 1391 01:09:47,500 --> 01:09:50,440 I mean, maybe this is polemical, but almost none of them 1392 01:09:50,440 --> 01:09:52,820 were designed for historical research. 1393 01:09:52,820 --> 01:09:54,860 So we take something that was designed 1394 01:09:54,860 --> 01:09:57,240 for a very different purpose, and we 1395 01:09:57,240 --> 01:09:59,780 think that we can now use this and apply it 1396 01:09:59,780 --> 01:10:03,020 to historical circumstances, contexts, documents, corpora, 1397 01:10:03,020 --> 01:10:04,620 and text, and so on. 1398 01:10:04,620 --> 01:10:07,420 And that's just too easy because it doesn't work. 1399 01:10:07,420 --> 01:10:10,740 So yeah, I think that having an unclear understanding, 1400 01:10:10,740 --> 01:10:14,500 critical understanding of tools, data sets, and sources 1401 01:10:14,500 --> 01:10:17,300 is eminently important. 1402 01:10:17,300 --> 01:10:18,020 Thank you. 1403 01:10:18,020 --> 01:10:18,960 More questions. 1404 01:10:18,960 --> 01:10:20,860 Daniel? 1405 01:10:20,860 --> 01:10:24,100 Yeah, thank you very much for your inspiring talk. 1406 01:10:24,100 --> 01:10:26,420 I've got a half question, half comment. 1407 01:10:26,420 --> 01:10:30,740 I would like to ask you to comment on my comment. 1408 01:10:30,740 --> 01:10:39,100 So you truly criticized or expressed 1409 01:10:39,100 --> 01:10:44,860 the importance of data literacy and data criticism, which 1410 01:10:44,860 --> 01:10:47,940 is to no offense, which is to some degree, 1411 01:10:47,940 --> 01:10:51,220 it's trivial because everyone who works with data 1412 01:10:51,220 --> 01:10:54,620 at some point gets to the point to discover that. 1413 01:10:54,620 --> 01:10:59,100 But my idea and my thought is data modeling 1414 01:10:59,100 --> 01:11:06,060 allows contradictory statements, allows multi-layered analysis. 1415 01:11:06,060 --> 01:11:13,220 We've got the tools and the ways of how to model data, 1416 01:11:13,220 --> 01:11:15,300 they are there already. 1417 01:11:15,300 --> 01:11:21,700 But I don't have an overview of all the digital history studies. 1418 01:11:21,700 --> 01:11:24,180 I'm not a historian myself, so I don't 1419 01:11:24,180 --> 01:11:30,380 have an idea, but what I know is that it's possible to-- 1420 01:11:30,380 --> 01:11:34,620 if you annotate in historical text, for instance, as you do, 1421 01:11:34,620 --> 01:11:38,660 and of course, there is always a step of interpretation. 1422 01:11:38,660 --> 01:11:44,980 But I would argue there's not an infinite numbers 1423 01:11:44,980 --> 01:11:49,620 how to integrate that, but more than one at least. 1424 01:11:49,620 --> 01:11:51,780 You could do that, it's much more effort, 1425 01:11:51,780 --> 01:11:56,260 but one could do that and compare 1426 01:11:56,260 --> 01:12:00,380 the different interpretations and the different annotations 1427 01:12:00,380 --> 01:12:05,060 and to even model contradictory narratives, 1428 01:12:05,060 --> 01:12:09,180 contradictory statements, analysis. 1429 01:12:09,180 --> 01:12:13,940 And I'm wondering why a part of-- 1430 01:12:13,940 --> 01:12:18,900 that's much more effort to do, but that's the chance 1431 01:12:18,900 --> 01:12:21,700 we have to write a new history. 1432 01:12:21,700 --> 01:12:24,700 And this kind of new history will not 1433 01:12:24,700 --> 01:12:26,460 be like a completely new history, 1434 01:12:26,460 --> 01:12:29,780 telling completely new things, but comparing 1435 01:12:29,780 --> 01:12:33,180 different perspectives, being more multi-dimensional. 1436 01:12:33,180 --> 01:12:37,820 And I'm just wondering why are we still on the step 1437 01:12:37,820 --> 01:12:39,100 before of that level? 1438 01:12:39,100 --> 01:12:40,380 I mean, people do. 1439 01:12:40,380 --> 01:12:43,620 And as Manuel Borchardt mentioned before, 1440 01:12:43,620 --> 01:12:47,820 people tend to use, let's say, the default parameters. 1441 01:12:47,820 --> 01:12:51,820 For example, you say, OK, people just bend things as they wish. 1442 01:12:51,820 --> 01:12:53,540 And I would argue other people just 1443 01:12:53,540 --> 01:12:56,860 use the default parameters, not being aware of what 1444 01:12:56,860 --> 01:13:00,220 would change the output, how the output would change. 1445 01:13:00,220 --> 01:13:04,980 So yeah, it was more like a comment than a question, 1446 01:13:04,980 --> 01:13:09,140 but maybe you could comment on this. 1447 01:13:09,140 --> 01:13:09,620 I can. 1448 01:13:09,620 --> 01:13:11,060 You can ask a question. 1449 01:13:11,060 --> 01:13:12,460 I can ask a question, no. 1450 01:13:12,460 --> 01:13:16,340 I totally agree with what you described as what is possible, 1451 01:13:16,340 --> 01:13:19,940 including the statement that it takes a long time 1452 01:13:19,940 --> 01:13:27,380 and it's a group effort, really, most of the time. 1453 01:13:27,380 --> 01:13:29,140 And I think this is where-- 1454 01:13:29,140 --> 01:13:31,140 so I totally agree with you. 1455 01:13:31,140 --> 01:13:33,860 And in a perfect world, I would always 1456 01:13:33,860 --> 01:13:36,100 try to create annotation schemes or sort of data 1457 01:13:36,100 --> 01:13:39,340 sets the way that you said, not only matching and making it 1458 01:13:39,340 --> 01:13:40,940 all look the same, like, for example, 1459 01:13:40,940 --> 01:13:43,780 the figuring out the past book kind of, in my view, 1460 01:13:43,780 --> 01:13:46,660 try to do, how do we describe societies today? 1461 01:13:46,660 --> 01:13:49,860 Let's take the same thing, describe past societies. 1462 01:13:49,860 --> 01:13:51,460 And then it gets shaky, right? 1463 01:13:51,460 --> 01:13:55,340 So you could integrate this, and you should integrate this. 1464 01:13:55,340 --> 01:13:57,860 But I'm a digital history professor 1465 01:13:57,860 --> 01:14:04,180 at a very normal German history department. 1466 01:14:04,180 --> 01:14:08,260 I do have a large group, because I've got lots of projects. 1467 01:14:08,260 --> 01:14:11,780 But I think that I still make this argument 1468 01:14:11,780 --> 01:14:14,660 towards other historians and also towards institutional 1469 01:14:14,660 --> 01:14:20,460 contacts to make sure that we need the resources to use 1470 01:14:20,460 --> 01:14:23,580 those possibilities. 1471 01:14:23,580 --> 01:14:26,740 And if you'd like, for example, the-- 1472 01:14:26,740 --> 01:14:33,580 I'm going to just show you the video, and now it doesn't work. 1473 01:14:33,580 --> 01:14:38,020 But basically, this project kind of came from this idea 1474 01:14:38,020 --> 01:14:45,420 to try to design not only another network graph, 1475 01:14:45,420 --> 01:14:48,100 but try to design a multi-perspective, 1476 01:14:48,100 --> 01:14:50,580 multilayered environment in which I 1477 01:14:50,580 --> 01:14:52,940 can interact with data points to paint 1478 01:14:52,940 --> 01:14:55,420 a picture of possibilities. 1479 01:14:55,420 --> 01:14:56,860 This took forever. 1480 01:14:56,860 --> 01:14:58,260 It was very expensive. 1481 01:14:58,260 --> 01:14:58,940 It is running. 1482 01:14:58,940 --> 01:15:00,100 So visit our web page. 1483 01:15:00,100 --> 01:15:02,940 You can actually use this on the screen. 1484 01:15:02,940 --> 01:15:05,980 Visit Bielefeld, and you can actually use this with VR 1485 01:15:05,980 --> 01:15:08,180 headsets and so on. 1486 01:15:08,180 --> 01:15:10,740 So long story short, I totally agree with you. 1487 01:15:10,740 --> 01:15:14,380 And I think that the reason why I still start at the point 1488 01:15:14,380 --> 01:15:17,060 where I'm starting is because I direct this 1489 01:15:17,060 --> 01:15:22,220 towards my own discipline to be more explorative and take 1490 01:15:22,220 --> 01:15:26,820 the opportunities that are there more openly 1491 01:15:26,820 --> 01:15:30,060 and give us the resources to do so. 1492 01:15:30,060 --> 01:15:31,300 All right, thank you. 1493 01:15:31,300 --> 01:15:33,700 I think we might have time for a very short last question. 1494 01:15:34,700 --> 01:15:37,340 OK, in a room full of historians, no short question. 1495 01:15:37,340 --> 01:15:37,860 Oh. 1496 01:15:37,860 --> 01:15:39,140 [LAUGHTER] 1497 01:15:39,140 --> 01:15:43,220 Read this short question, I'll be back to you. 1498 01:15:43,220 --> 01:15:49,300 OK, yeah, getting back to what Julius asked about the-- 1499 01:15:49,300 --> 01:15:50,100 yeah, what is data? 1500 01:15:50,100 --> 01:15:52,060 What are the sources? 1501 01:15:52,060 --> 01:15:53,820 And I found it interesting what you said. 1502 01:15:53,820 --> 01:15:58,460 So data is what you create yourself, more or less. 1503 01:15:58,460 --> 01:16:01,060 And the other things are there also. 1504 01:16:01,060 --> 01:16:07,780 And I don't know because the background is that also in our 1505 01:16:07,780 --> 01:16:10,820 SFP, we discussed at length what are the data, actually, 1506 01:16:10,820 --> 01:16:12,420 that we're supposed to manage and so on, 1507 01:16:12,420 --> 01:16:14,500 and how is it different from the sources and material. 1508 01:16:14,500 --> 01:16:17,300 And I think what the perspective or the definition 1509 01:16:17,300 --> 01:16:20,260 you provided now gives then the impression. 1510 01:16:20,260 --> 01:16:22,820 But of course, also the sources are not just there. 1511 01:16:22,820 --> 01:16:25,860 Someone did something to them before, put them, selected 1512 01:16:25,860 --> 01:16:26,900 them, and so on. 1513 01:16:26,900 --> 01:16:28,900 And at the same time, if you want to do that, 1514 01:16:28,900 --> 01:16:31,140 put them, and so on, and at the same time, 1515 01:16:31,140 --> 01:16:34,820 if you work with the data from others, it's also-- 1516 01:16:34,820 --> 01:16:37,220 then we call it data, but also someone else created it. 1517 01:16:37,220 --> 01:16:39,300 So I'm not sure if I-- 1518 01:16:39,300 --> 01:16:43,700 yeah, may I also find your distinction a bit difficult? 1519 01:16:43,700 --> 01:16:45,700 I understand it, but-- 1520 01:16:45,700 --> 01:16:48,980 and of course, it's not just a question about denomination 1521 01:16:48,980 --> 01:16:50,420 and how do you define things. 1522 01:16:50,420 --> 01:16:51,860 But then also to get to the point, 1523 01:16:51,860 --> 01:16:55,380 what is different now with the digital is-- 1524 01:16:55,380 --> 01:16:57,860 I don't know, talking also in the NFT for memory, 1525 01:16:57,860 --> 01:17:00,900 how is it different with the digital source criticism now? 1526 01:17:00,900 --> 01:17:04,100 How is it different from the traditional source criticism? 1527 01:17:04,100 --> 01:17:06,020 So I don't know. 1528 01:17:06,020 --> 01:17:08,980 It's also not really a question, maybe a comment. 1529 01:17:08,980 --> 01:17:10,740 Yeah, maybe you want to comment. 1530 01:17:10,740 --> 01:17:13,460 Like to your last point, what are the differences? 1531 01:17:13,460 --> 01:17:16,580 I think this is really not trivial. 1532 01:17:16,580 --> 01:17:18,820 Because I think that in the world that we live in 1533 01:17:18,820 --> 01:17:21,620 and with the tools that we have at our disposal, 1534 01:17:21,620 --> 01:17:24,500 lots of things can be called digital humanities 1535 01:17:24,500 --> 01:17:27,780 or digital history already. 1536 01:17:27,780 --> 01:17:32,980 If you read something, print it out, or like a book, 1537 01:17:32,980 --> 01:17:35,540 and you use your pencil to make notes, 1538 01:17:35,540 --> 01:17:39,860 or you use this on a tablet and you use a pencil to make notes, 1539 01:17:39,860 --> 01:17:41,940 that's not really a real difference, right? 1540 01:17:41,940 --> 01:17:44,140 So you shift the medium, OK, check. 1541 01:17:44,140 --> 01:17:46,460 And there are studies that reading digitally 1542 01:17:46,460 --> 01:17:50,660 is different to reading books, all that taken into account. 1543 01:17:50,660 --> 01:17:53,900 But it doesn't really change our practices so much, 1544 01:17:53,900 --> 01:17:56,780 because we still read and annotate. 1545 01:17:56,780 --> 01:17:57,980 And I think it is a really-- 1546 01:17:57,980 --> 01:17:59,420 and I'm not answering the question, 1547 01:17:59,420 --> 01:18:00,540 but I think it's really important 1548 01:18:00,540 --> 01:18:03,940 to stick with that question and see which kinds of practices 1549 01:18:03,940 --> 01:18:06,980 do we just transfer into the digital, 1550 01:18:06,980 --> 01:18:08,380 and where does the digital actually 1551 01:18:08,380 --> 01:18:10,740 transform the practice? 1552 01:18:10,740 --> 01:18:15,420 And I think this has not been sufficiently answered. 1553 01:18:15,420 --> 01:18:18,340 And I think this is also why it's not only the digital, 1554 01:18:18,340 --> 01:18:21,220 but it's the specific computational methods. 1555 01:18:21,220 --> 01:18:26,540 It's sort of the opening of possibility spaces like VR 1556 01:18:26,540 --> 01:18:29,180 or whatever you have, where you can really 1557 01:18:29,180 --> 01:18:32,140 explore the transformations and see, OK, 1558 01:18:32,140 --> 01:18:35,060 what is it that we could not have done before? 1559 01:18:35,060 --> 01:18:37,140 And I also don't like that the argument often 1560 01:18:37,140 --> 01:18:40,780 is that I can now have 4,000 sources instead of 100. 1561 01:18:40,780 --> 01:18:42,860 Yeah, OK, so that's scaling. 1562 01:18:42,860 --> 01:18:44,620 Fine, but is that a transformation? 1563 01:18:44,620 --> 01:18:47,900 Does that really change the way that we think about stories? 1564 01:18:47,900 --> 01:18:49,620 And that was sort of part of the argument 1565 01:18:49,620 --> 01:18:51,980 that I wanted to make, that I think 1566 01:18:51,980 --> 01:18:53,820 that we have the potential to actually think 1567 01:18:53,820 --> 01:18:55,420 about stories in a different way, 1568 01:18:55,420 --> 01:18:59,700 and also maybe use this along the lines of digital history 1569 01:18:59,700 --> 01:19:01,780 to also teach people more about perspectives, 1570 01:19:01,780 --> 01:19:05,460 use these kinds of methods not only for our own research, 1571 01:19:05,460 --> 01:19:09,740 but use it to teach people about history and the perspectiveness 1572 01:19:09,740 --> 01:19:12,060 of an historical argument. 1573 01:19:12,060 --> 01:19:15,180 And I think the source and the data thing, yeah, 1574 01:19:15,180 --> 01:19:17,580 maybe data sets should then be treated as sources. 1575 01:19:17,580 --> 01:19:21,660 And we need sort of rules for source criticism for data sets. 1576 01:19:21,660 --> 01:19:25,900 In this regard, data is also a source. 1577 01:19:25,900 --> 01:19:27,320 But the main argument that I wanted 1578 01:19:27,320 --> 01:19:29,140 to make towards also Julius's question 1579 01:19:29,140 --> 01:19:31,420 is that, for me, it's too easy. 1580 01:19:31,420 --> 01:19:35,700 If I ask people with what kinds of data do they work, 1581 01:19:35,700 --> 01:19:38,260 and they sort of show me a picture of a medieval 1582 01:19:38,260 --> 01:19:42,060 manuscript, then I need something 1583 01:19:42,060 --> 01:19:45,100 to start a conversation, right? 1584 01:19:45,100 --> 01:19:46,500 So I think it's important to make 1585 01:19:46,500 --> 01:19:49,740 the differentiation between a manuscript and the data set, 1586 01:19:49,740 --> 01:19:50,860 but both can be sources. 1587 01:19:50,860 --> 01:19:56,380 Maybe that's more correct to sort of have a look at this. 1588 01:19:56,380 --> 01:19:57,060 All right. 1589 01:19:57,060 --> 01:20:00,140 I think that's the end of the official discussion part. 1590 01:20:00,140 --> 01:20:01,780 Thank you so much, Silke, for this talk. 1591 01:20:01,780 --> 01:20:03,020 Thank you for your question. 1592 01:20:03,020 --> 01:20:06,380 [APPLAUSE] 1593 01:20:06,380 --> 01:20:09,440 [MUSIC PLAYING] 1594 01:20:09,440 --> 01:20:12,480 [MUSIC PLAYING] 1595 01:20:12,480 --> 01:20:15,060 (bright music) 1596 01:20:15,060 --> 01:20:20,060 ♪ ♪ ♪