 We should be live as usual. Wait for a few moments, okay? And everyone is, should be pulled into this session now. So welcome everyone to the last two talk block of the meeting, or not of the meeting, of the day. Sorry, the day's, it's only day three. I'm very, very excited to present our next speaker. So one thing that I think is important, I was hoping in this meeting to take seriously the interaction between digital studies of science and studies of digital science. And digital studies of science doesn't have to mean just things that have been done since 1995 or something. So I'm very excited to have some proper history of science on the program. So it's my distinct pleasure to introduce Jeffan Hesbrug and Walter with co-author work with York Walter who can't be here on digital interdisciplinarity in dissertations in the Holy Roman Empire. So please, without further ado, take it away. Thank you, Charles. It seems to be a tourism that there is no interdisciplinary without disciplines. And the received wisdom in the literature on the history of interdisciplinarity contends that disciplines were established only in the course of the 18th century. So if we are interested in how interdisciplinarity evolved, we must investigate the history of knowledge starting from that point. This assumption is problematic. I acknowledge that the term discipline comprises more than a rubric and a classification of knowledge as we find it, for example, in the polygamy now of late ancient commentaries on the Aristotelian corpus or Augustine's late octrina cristiana or later in the encyclopedic projects of thinkers like Alstead, Comenius, or Leibniz. But disciplines like history, philology, logic, or natural philosophy did have effects on the organization of institutions of higher learning in the period I'm interested in the 17th century. They served as denominations for professorships. They shaped the organization of books and libraries. And, and this is the topic of today's talk, they were used as rubrics for the classification of early modern dissertations. Whether or not this suffices to talk about the existence of scientific disciplines in the 17th century in the full sense is maybe an open question. If you feel uncomfortable, you can also substitute my talk of disciplines with talk of sub-disciplines or proto-disciplines. I won't go into a bad discussion at large. In this non-butted project I present here today, we look at various ways to process metadata for early modern dissertations published between 1601 and 1700 as they are recorded in Germany's National Bibliography for this period, Faudi Zipsen, VD 17. Today I want to present work on a subset of these metadata comprising more than 900 interdisciplinary dissertations, dissertations that contain more than one term for discipline or sub-discipline or proto-discipline in their title. They make up roughly 4% of the total copies. Such an inquiry is itself interdisciplinary. We operate based on the premise that at least for the 17th century, history of science and history of the humanities should cooperate in the investigation of what one could call history of scholarship. And for this we use the work products of professionals in information science, namely librarians, specialized in cataloguing all the prints. And the methods we use to trace the evolution of interdisciplinarity in our data set and to identify interesting patterns are themselves interdisciplinary, namely those of the digital humanities. Interdisciplinary dissertations first came to our attention during our attempts to classify the dissertations in question according to the sub-discipline they belong to. For this, and this was the work done by Jörg Walter, we use machine learning algorithms for which the correct classification of interdisciplinary dissertations was a significant challenge. So we decided to investigate these titles separately. The second factor to be mentioned is the regrettable lack of uniformity in the capture of titles in the original data set. We can't discuss details here, but originally only 14,650 titles were amenable to machine learning-based disciplinary classification. While the remaining 5,800 contained additional title elements that created too much noise for machine learning and algorithms to succeed. Our interdisciplinary data set contains 629 interdisciplinary titles in the classifiable data and 330 in the non-classifiable data set, taking into account stop words like toponyms and so on, we end up with 920 dissertations. And that's a preliminary number, please don't quote it. This is still work in progress. If we now look at how the labels for such dissertations are distributed, the largest number are dissertations in the intersection of history and politics marked as historic politic. I've simply taken the stamped forms we have used to search to label the graph. The second position, historical philological dissertations, but historical or philological expertise can also be coupled with other philosophical disciplines, for example, physics. So we also find 32 historical physical dissertations. And another coupling that is worth mentioning, 22 logical metaphysical titles. And I confess I'm a bit envious listening to all these talks with thousands of papers in the corpus. We are talking here about dissertations published over the course of a century, and then 22 is not that much. So don't put too much trust in absolute numbers or even percentages. The conclusions we can draw here are largely qualitative. The preponderance of historical and philological disciplines is of course also mirrored when we look at disciplinary labels across interdisciplinary titles and count them separately. History is leading with 500 mentions. The next disciplines are politics or political philosophy, philology and physics, by and large in agreement with the image regard from the counting of the interdisciplinary labels as such. This bar plot also shows the relative proportions of labels in the first position. For example, historical in historical politica, that is blue. And the second position of these compounds, politica in historical politica, orange. This finding suggested that there might be some difference between both positions so that we could identify one position in these compounds as superordinate and the other as subordinate. But analysis just of the interdisciplinary titles as such showed that between the two spheres within each disciplinary label, there's almost no overlap. So we had within the titles as such no real indication for how the two halves of these interdisciplinary compound, adjective compounds relate to each other. But I return to that problem later on. We also looked at the temporal evolution of the subgenre. There you can see that the proportion of interdisciplinary dissertations in relation to our data set of philosophical dissertations in total increases throughout the 17th century. Again, don't put too much trust in the linear regression. But I think even if you just look at the scatter plot as such, the increase is more or less. Obvious, so this shows just to clarify how many percent of philosophical dissertations in a given year were dissertations with an interdisciplinary label. And you see some high points around 10%. In other years, it was around 1%. So there's a certain variability. But again, these numbers can be interpreted qualitatively, but the record we have has serious holds. So this should be taken with a certain amount of caution. If you then look whether we can identify authors which put who put a special emphasis on interdisciplinary dissertations. What we find is that the absolute count of interdisciplinary dissertations as such is not a sufficient indicator for relevance. The example for that is the last person in this list. Jakob Tomasius, at least colleagues who have some acquaintance with the history of early modern philosophy may have heard the name as a teacher in correspondent of Leibniz. And we have in German libraries eight interdisciplinary dissertations, which is not, yeah, which is quite considerable number, or also if you compare it to the other authors in this table. But we must put this in relation to the overall output from Tomasius that we can find. And those are 152 dissertations. So only 5.3% of his overall output were interdisciplinary, while for the other four authors, the percentage is much higher. So if you want to know more about the background of these interdisciplinary dissertations, you should look at those authors for which they formed a considerable part of their output, not just at the absolute numbers. And those are names that even most specialists on the history of German philosophy in the 17th century may not have heard. Maybe Agidius Strauch, who was a student and professor in Wittenberg and later Rektor in Danzig, he published mainly dissertations on historical sub-disciplines like geography. For example, a dissertation on Iceland with a respondent coming from Iceland, a co-author. Matthias Berneger was a correspondent of Galilei and Kepler, a professor in Strasbourg. And he's present in our corpus mainly with historical political dissertations, for example, on the migration of students or the kingdom of Hungary. Roth was a teacher in Ulm, wrote historical political and historical philological dissertations on the history of torture, but also maybe to compound for his sins, historical ethical dissertation. And Hopfer was a professor in Tübingen who wrote historical physical dissertations, for example, on the spontaneous generation of animals living in fire, the salamander and the pyrausta, I think. Yeah, this is just to give you some cursory impressions what kind of material I'm talking about. What we have looked at in some depth are what you could call topics of dissertations. And that's more easily explained using a typical dissertation title. This is one I've made up, so you don't find that in the corpus. Let's assume someone published a dissertation, historical physical Dinnatura Acre, a historical physical dissertation on the nature of water. The biogram we are interested in that signifies what we call the topic would be Dinnatura, the essence of being or the nature of water. And what we wanted to know is, does Natura appear in historical dissertations, in physical dissertations, in both disciplines making up the compound historical physical or in none? And these are the results, 72% of our corpus have no biogram that is used in either historical or physical dissertation. 21.9% of our corpus use the topic that is mentioned in the dissertation in one of the constituent disciplines, so either historical or physical, logical or metaphysical. And 72 dissertations have a topic that appears in both of their constituent disciplines. We now take a closer look at the distribution among disciplines of these one topic dissertation titles. So in 36.09% of historical political dissertations, the biogram starting with D occurs in either historical or political dissertation. I won't read out the whole table. You see that the numbers vary age, especially low in logical metaphysical dissertations high in historical political. This is just to show that averages across the whole corpus may obliterate differences between disciplines. And now we come back to the question to which extent may the position of a discipline label in this interdisciplinary compound help us to determine whether one discipline is regarded as primary in contrast to the other. And for this we looked at these topical biograms. Those biograms appearing only in one discipline, those biographs appearing in two disciplines don't help us to decide which discipline is more important and those that don't appear in a discipline at all don't have either. And there we found a certain uniformity in that disciplines appearing in the first position, historical, etical, philological share less topics with dissertations in this you could say unified, unitary discipline. So only 20% of topics of dissertations with philologic core also appear in purely philological dissertations. 80% of topics where historical is in the second position of the interdisciplinary compound have a topic that also appears in historical dissertations. To cut a long story short, it seems that those dissertations that obey our criterion, topic only in one disciplinary class are largely subsumed under the adjective in the second position. So historical political dissertations are primarily political, historical philological dissertations are primarily philological, historical, physical dissertations are primarily physical. So what we could not solve through an internal analysis of the titles could be solved when looking at the presence of what we call topics in purely disciplinary dissertations. This already leads me to my discussion and conclusions. Our data show that most interdisciplinary dissertations were produced in the overlapping area between the nascent disciplines of history and philology on the one hand and disciplines of practical philosophy, most prominently politics on the other hand. The largest interdisciplinary group of titles with at least one theoretical subdiscipline of philosophy are historical physical dissertations who came in at rank four with 32 titles. We also saw an upward trend regarding the overall number of interdisciplinary dissertations in the second half of the 17th century. And I have the hunch that these two developments on the other hand the prominence of history and philology and on the other hand the increase over the 17th century link these interdisciplinary dissertations to a phenomenon that historians of knowledge in Germany in this period call polyhistoricism or polyhistoria, the appropriation of knowledge from many different domains, not because people were looking for generalized insights into the essence of the world, like in evolving natural science or in metaphysics or in Aristotelian natural philosophy. But with an interest in individual phenomena such the original meaning of history, so polyhistoria, knowledge of many different phenomena are finding that for 70% we either don't have a topic or sorry, the number was 40% have a topic that does not appear in disciplinary dissertations may further bolster this thesis. We will now have to look how the supervisors, the presides of these dissertations figure in this broader development in what I call the history of scholarship precisely because this is not purely history of science, this is not purely history of the humanities. Our analysis of titles has also shown that it might make sense to distinguish two senses or modes of interdisciplinary in this text, what one could call object-bound interdisciplinary which is based on a topic that is relevant for both disciplines and appears in single discipline dissertations in both disciplines for those who have background in the history of philosophy in this period, paradigmatic cases for these are dissertations on the soul, disputed area between natural philosophy and metaphysics and logical metaphysical dissertations, for example, on the categories, again a topic that is discussed both in logic and metaphysics. And from this we may distinguish method-bound interdisciplinary and that might be connected with those dissertations where the topic is present only in one discipline. So if you write a historical physical dissertation on Natura, Natura is a term that appears predominantly in physical dissertations and is absent in historical. What does history add? History may add an additional methodological perspective. Finally, we have also seen that if there's only one topic and a topic for one discipline in the dissertation title in the majority of cases across various interdisciplinary, this topic is mostly found in disciplinary dissertations for the discipline in the second position. So this means that in the majority of cases, though not exclusively, we can surmise that these compounds like historical physical, historical, political are as you would say in linguistics, hypotactic, subordinating the first to the second discipline. And that's it already. What we can learn from metadata may be to a certain extent limited, but with a certain amount of creativity and ingenuity, you can force the water to come out of the stone. Thanks so much. That's exactly what I was gonna say. It's impressive how far you can go on the metadata. This is really interesting. While I wait on the chat to catch up, I have a bunch of questions of my own, but let me start with, let me start with perhaps this one, which is more clarificatory. So in this period, what is the role of a dissertation in the career of the person who writes it? So I'm thinking that some of the touchstones that I have for this are radically different. That is to say, the supervisor would sometimes write the majority of the dissertation or something like this, right? So I'm wondering for this period in Germany, what's the life of this object for these people? Yeah, that would be a separate presentation because I've also engaged in a cosmographical comparative analysis of presides and respondents. And for example, if you look at the age curve, the age of presides, what we call today supervisors, things rapidly after 30. So for a lot of authors of these dissertations, they presided but presided as a qualification, namely as a qualification to be allowed to teach philosophy. But as it is today, a lot of people after that left the profession in the early modern university, not necessarily academia, but they tried to advance to theology and law and medicine, not least because those were better paid positions. So after 30, you only find dissertations where professors really supervised. The question of authorship is something I studiously avoid because reams of pages have filled with this without any ultimate conclusion. In some cases, people indicate that even though they appear as respondent, it is them who have written the text. But this goes back to these problems about the capture of titles I mentioned in the beginning. The bibliography we are using does not always record everything that is on the title page because they also have reproductions of the title page. But the way the material on the title page is reproduced is not uniform either. So different schools of librarians over the 20 years of this project chose different things to include into what we can get as machine readable data. And that's a huge deficiency, but I don't want to complain too much because people planning this project in the 1990s could not anticipate the rapid evolution of various methods we could use to process this data. Sure, yeah, that makes very good sense. Nicola Bertoli in the chat asks or says interest, first of all, starts by saying interesting talk and asks, so I was wondering whether your corpus reflects the evolution of the meaning of the term natural history, like Historia naturalis in the early modern period or if there's other ways that you've caught, now I'm expanding on this question, changes in this terminology over the period of the century of metadata. Yeah, the deacronic analysis, especially of these interdisciplinary titles is something still on the agenda. What I can say is that if I have messed up the regular expression in searching for these interdisciplinary terms, there are no dissertations in natural history. And this strengthens my claim because they belong to historico-physica. And this strengthens my claim that these are disciplinary or terms for disciplines or proto-disciplines because you had to pick an area which was in some way socially effective within the university. For example, as the denomination of a professorship and there are no professorships for natural history. So there cannot be dissertations that have the interdisciplinary label Historia naturalis, historical naturalis or natural historical or whatever you may make up there. And since the number of titles, so I will look at the diacronic distribution, but I have not much hope because 900 sounds okay, but 900 over 100 years, it's just 10 per year. And I anticipate that the data will be so noisy that you cannot make really well-founded assumptions. And additional problem, dissertation prints were an extremely ephemeral genre. So we can estimate that we have lost more than we have preserved. And maybe it's only a quarter or 20% or even less that we find today in libraries. And there are also certain geographical biases in the bibliography because certain areas in Germany did not participate in this, like the West and the South. So I'm a bit careful, I have not that high expectations that we can really get to a fine-grained historical analysis also disciplines evolved over the whole century. Sure, no, that makes sense, that makes sense. Monimus Rahi has a pair of questions. So first, similar to the previous about natural philosophy, about that other way of picking up the same kind of point. Actually, because we have plenty of time, let me go ahead and let you answer, respond to that really quickly. It may be a similar answer, so. Yeah, that would be physical. And as I said, we started, this is maybe something I should explain. We started to look at the interdisciplinary dissertations for two reasons. For one, Charles, you and I had a conversation on Twitter where you complained that there is not much work on the history of interdisciplinarity. And at the same time, I looked at our results and found out that it's especially these interdisciplinary dissertations which create additional noise that ruins our precision and recall. So I thought we could put them apart and look at them separately, but what is still missing is now to rerun the classification of the original dissertations without the noisy interdisciplinary data. And after that, I will be in a better position to answer questions regarding, yeah, what are they called, unitary disciplines? If you want to contrast a discipline and an interdiscipline, what's the end of the discipline? Yeah, I don't know, but fair. Yeah, yeah, yeah. And let me pick up with the second half of this question. So also, could you give it an example of this method bound interdisciplinarity? So for example, like, for instance, when like Hume is trying to introduce the experimental method of reasoning into moral subjects, is that the kind of thing that we have in mind here? No, it's rather, I mean, I have looked only at the titles and the titles themselves tell you what the dissertation is about, but they don't tell you the content as such. And I believe in the context I'm working in, what the age can achieve is to structure a field in ways that are less prejudiced than our traditional scholarship in this area. For example, a lot of work is done on school metaphysics and school logic because this has effects on the history of philosophy as we perceive it in the 18th century. So prehistory, pre-prehistory of Kant, for example. And the history of natural philosophy in Germany in this period is comparatively under-researched, although it is a more prominent discipline, presumably, than metaphysics and logic. But as long as we have no reliable way to OCR this stuff in a way that makes the full text available, there is a wall which digital methods cannot get over. So as for now, this would indicate which of these dissertations carry a special interest for a researcher interested in this period as a whole. And the argument for this is not just the reading experience and the intuition of the scholar, but objective facts of how this or that author or group of dissertation fits into the overall landscape. So it's distant work without the reading. I like it, I like it, that's a great way to put it. Question from Rose Trappes who asks, maybe I missed this in the talk, it's been a long day, but I'd be interested to hear if you could speculate for the reasons for the dominance of those disciplines in the interdisciplinary dissertations that were so, that were so predominant or is it too hard to say without being able to have access to the content? Yeah, if my assumption that this is linked to the development of polyhistoricism is correct, then it was historical and philological disciplines with their, you could say, idiographical approach. So not interested in laws, but in the individual phenomenon that were the drivers of this broader polyhistoric movement, you could say. And then if the dissertations are in fact connected to this movement, then historians of this part of the history of knowledge will readily agree that it's then no surprise that the majority of dissertations are in some way linked to history and philology. And the increase over the 17th century may then be due to the increased establishment of these disciplines within the philosophical faculty. So in the beginning of the 17th century, you may have people doing work in philology, but maybe not that many in Germany and certainly not as professors for their discipline, but the institutionalization of philology and history as university disciplines within the context of philosophy took place in the second half of the 17th century. Very interesting. Okay, so that's helpful for a clarification, well, a sort of clarification and expansion question that I wanted to ask. I'll pose it anyway to see if there's anything you want to add about it, but that was something that struck me was this upward, this upward pointing trend, this pretty remarkable upward pointing trend over the course of the century. And I guess, I don't know, there's just a strange way to phrase this, but just like, is this, does that make sense with our understanding of where this trend would go in the 18th, just to the extent that we have an understanding of where that trend would go? Were you expecting that? How to, cause that's a pretty, I mean, that graph is pretty striking. Again, under the assumption that my hypothesis about polyhistory is not completely misguided, we must know two things. First, at the beginning of the 17th century in the early enlightenment, it was exactly this kind of, do people know the German word Wunderkammer? Like these rooms full of strange, miraculous objects and walk around. I learned about it in my HPS graduate training, but I'm not sure how many people in the audience have HPS. Yeah, so these were basically proto-museums at an aristocratic court where people collected lots of many different strange things from all over the world and from all ages. When I think about polyhistory, I think about this kind of arrangement of objects and knowledge about objects. And this came under heavy scrutiny in the early enlightenment. So being a polyhistor in the 18th century was not something you could be proud of because the early German enlightenment emphasized exactly the opposite, the importance of thinking for yourself. Self-sting, autonomy and polyhistory is the exact opposite. This is also, of course, an explanation of why this phenomenon is not widely known because people have bought this bad press, enlightenment authors, the bad image enlightenment authors created. Besides that, more concrete trends will have to wait until the project for the artsen. The subsequent bibliography for the 18th century is finished. It's underway and you probably can get data, but they are not as complete as I would like them to be. So this is what I'm waiting for. I see. So, okay. So actually that's a question that had occurred to me. So the current status of the project is they have, the 17th that you've done is completely finished, but the 18th isn't yet. Yeah, exactly. Okay. With that, we're very close to time. So I think what I'll go ahead and do is wrap it there. Thank you one more time. Thanks so much. This was a really, really neat talk. I really enjoyed getting to see this. And with that, yes, the last talk of the day will be back in just a few minutes. So thanks so much. Thank you all. Bye.