 I'm David Rowe, I'm the Director of Digital Matters and by honor to introduce Jeremy Brown, who is a Associate Research Professor at Brigham M. University in the Office of Digital Committees. Today he's going to be talking about the Women's Exponent Advertisement Database, Jeremy Brown. Thank you, David, and thank you for having me. I mean, how far back do we go? The minute I step into Utah, you're costed. We've been collaborating on various things for years. It's been a lot of fun, and it's great to see Rebecca and Anna here as well, some of my friends. Those of you online, welcome as well. I want to go over a little bit of the format of the presentation today because I really want this to be interactive. I've ordered what we're going to go over in terms of what was advertised. We can spend the whole time talking about the finished product and then have a little bit of time to talk about the technology behind it, and then you'll fully forget about this commentary that I have at the end. That's not important. If you have questions about the database or about its uses or some of the research we've done with it, or if you have ideas, please go ahead and stop, raise your hand or give a comment to Comstock, and we'll make sure to involve everyone here because I really do want this to be interactive. Who am I? What do I do? As David said, I'm Jeremy Brown. I'm an Associate Research Professor in the Office of Digital Humanities at BYU. Before coming to BYU in 2012, I was at SUNY Brockport, which is near Rochester, New York, and then this is what I do, or more specifically what I have done at BYU. These are a collection of most of the projects that I've worked on while I was there. That 3D printed Cuneiform tablets. That was a fun one for TJ who does the 3D printing here, and some of these I've been in collaboration with people here at BYU and others with other people. I'll bring this back at the end though. So let's talk about the exponent advertisement database. The finished product is a website where you can search 4,000 advertisements that appeared in the women's exponent newspaper between 1872 and 1914, which is the entire run of the newspaper in Salt Lake City. There are 112 vendors and 12 industries that I just coded myself. I would like to go back and redo that with like a Euro labor statistics kind of taxonomy instead of just one that I threw together. But let's take a look at it. There it is. All right. So I hope the screen sharing is working and following this. But you have just a simple search field where you can limit the years that you're interested in. You have specific industries that you can select. You can browse by year, industry or vendor. You can make some pretty simple graphs which we'll get to in a little bit that just show the frequency of advertisements by industry. There's an about page which includes acknowledgement to digital matters and the Willard J. Marriott Library, University of Utah. And so because this project grew out of the effort in 2019 to re-digitize the women's exponent newspaper in time for the sesquicentennial of women voting in Utah. Those celebrations were kind of curtailed by the pandemic unfortunately. But I saw very early that there was some potential for using these advertisements or targeting these advertisements for curation. And so Rebecca and Anna and Jeff and Jeremy and probably missing other people, David, who Elizabeth, who I'll mention later, I call him Brian Croxall at BYU, they were all involved in this re-digitization project. And that's where this came from. So there we go, we're back. Okay. Oh, by the way, I failed to mention that because ads were reprinted over and over again, there are 328 unique advertisements in them but a total of 4,000 printings of advertisements. All right, so what are some uses of this? Well, digital humanities, we got to make a graph, right? So this is the total number of advertisements that appeared by year in the newspaper. Sorry, I didn't know, you guys could see that too. And one of the ideas that we had early on was just to kind of use this as a surrogate for the economic viability of the newspaper and try to decide whether or not these peaks and valleys coincided with any major political or social events of the time period. The two big ones being the Edmunds Tucker Act that illegalized polygamy and strips women in Utah have the right to vote. And the Utah Statehood, really, there does not appear to be any sort of correlation to these changes. We don't know why the advertisements peaked in 1878 and dropped off precipitously. That would be a subject for investigation, but can't really tell that from here. By the way, this chart is the, there we go, is this graph right here that the website can make for you as well. And then also if we look by industry, say for example, the general stores, you can see that traditional retail was very prominent in the newspaper up until about 1888, and then it just dropped off the charts. Travel, the Intercontinental Railroad was completed over here, but we really don't get a prominent appearance of railroad advertising until the late 1880s and into the early 1900s. And then finally banking as a similar thing like this, these are entirely advertisements for Zion's Bank entirely and that will come back up in another slide as well. So those are some of the things you can do with a database like this, but I'm more interested in using this data to support and carry out research. So at Mormon History Association this summer, Elizabeth Smart from B. W. A. S. Library, she's our Digital Humanities Librarian and teaches in our Women's Studies Program. She and I presented at MHA and this was kind of a proof of concept for what you could do with these kinds of this kind of information. And I like to think of looking through advertisements as kind of digging through historical garbage, because nobody makes an advertisement with the idea that it's going to be preserved and reviewed in 50 or 100 or 150 years. They're making an advertisement for a specific purpose at a specific time. And so there's less of a filter, I want to say. It establishes more of a perception of economic reality than what we might get from an editorial. So what Elizabeth did that was so brilliant was she looked for editorials critiquing women's fashion. So you have to understand that the women's exponent was the women's newspaper for the Church of Jesus Christ of Latter-day Saints in Salt Lake City. And some notable names like M. L. I. D. Wells were editors. And there were several times editorials critiquing the worldly fashions of the day, particularly taking aim at French fashions and corsets and things like that. And what was interesting was to see the way that advertising would mention fashion, and that when the first editorial against those kind of fashions appeared, this advertisement, which had been there from the beginning, Mrs. Wilkinson's bizarre fashion disappeared in that issue. It did come back at least six months later, but other places continued to advertise for these kinds of fashions. There was even a French hair store that was advertising throughout the period where these critiques were being made. So there seems to be some dissonance between, or perhaps the advertising proves the need for these editorials. That people were buying these things, but there does not appear to have been an editorial privilege over the advertising, because there doesn't, except for that one case, there doesn't appear to be much reaction between them. But this is the kind of advertising that I really like, or kind of research that I really like, because it's looking at sort of the internal consistency of the newspaper and finding violations of that internal consistency through the advertising that appeared. In terms of research support, it was fun walking around MHA and seeing a poster for, I believe it was Ellis Schipp. There was a graduate student who does research on Ellis Schipp, who was an early woman physician in Utah. And I actually have scores of advertisers that, for example, this one here is from Romania Prat, who was another female physician, but it states that she is upstairs in the old Constitution building. And we have announcements in the advertisements of when they moved their office to another building. We have their telephone numbers sometimes, their early three-digit telephone numbers. So we can support research in that way. If you're doing any sort of research on a historical person who might have taken up advertising at this time, you can search the database and get more information on. So I think the second one is Zyna Young. Zyna, Zyna, I heard it both ways. One of Brigham Young's wives, who was very involved in the silk trade in Utah. And this is an advertisement that she's selling silkworm eggs from her residence, $4 per ounce. And then finally, let's take a look. If I go to search, and I just search for sewing machines, you can see, you know, weed sewing machines. Down here you have the premium distribution. Oh wait, no, so this is interesting. The reason this came up is because this is actually the Salt Lake Semi-weekly Herald giving away prizes for subscriptions, including a sewing machine. So you can you can win $100 down to $5 valuable foreign stock. So cattle, you can win from subscribing to the semi-weekly. But those are just some of the things that you can do, or that you can search to support resources be very much a secondary resource for historical research. This one I really like because it's one of those moments when you're reading something and it hits you. Like you come to realize the depth of something that was never meant to be deep. But for 10 years, Zion's Bank, which was a, you know, is a, I don't know if it still is, but it was a wholly owned subsidiary of the Church of Jesus Christ of Latter-day Saints. You can see Wilford Woodruff, who was the fourth president of the church, is the president of the bank here as well. But for 10 years, Zion's Savings Bank and Trust ran this advertisement that included this verbiage. The laws of Utah permit married women and also children who are minors to open savings account in their own name subject to their own order, meaning their own control. Have you such an account? If not, open one now. And my collaborator, Elizabeth Sparks, she would point out that because there was so much home industry going on at this time, this was not just a call for women to take the allowance their husbands were giving them and put them into the bank. That there was serious economic activity happening between women and amongst the female population at the time. But for me, it's this idea that the bank has to inform the readers of what the laws of the land allow them to do. And we've done some further research on this, and things that I didn't know from my position of privilege were that in the United States, women were required to receive their husband's permission to get a credit card into the services, and that there wasn't laws protecting them in that way. But it's just that the fact that the bank would need to state this for 10 years and their advertisements, really it's one of those insightful things that you get from digging through the trash. All right, another one that I like is the railway systems published fairs. They're scheduled to different places in Utah, including places that no longer exist, such as Frisco, which I had no idea was there. So I got on Wikipedia and found this picture of a ghost town. You can still visit it today, but it was a railway stop for the Union Pacific River. I can't remember which one it was, and I didn't have time to look it up, but there are places on here that we don't know where they are now. They were towns somewhere between these two towns because they're on the train schedule, but they're totally gone. There's no remnant of them. So this is again a little piece of historical information that may not be preserved in many other places. So and then we come to the oddities, and these are the funnest things usually to look at because we can laugh from our perch of enlightenment back on the past. The American Institute of Phrenology advertising their session to open September 3, 1902, or the Viabi treatment, which was a treatment for female ailments, which I did find some contemporary medical associations rebuttal. Let's put it that way of their methods. It was pretty much snake oil. And then down here at the bottom, Mrs. HK Painter and the electric physician electricity administered when cases require. So that was kind of interesting as well. And then this was my favorite oddity found. And this one is odd because, well, you have to know quite a bit of Mormon history to understand, but the Church of Jesus Christ of Latter-day Saints traces its origins to Joseph Smith, who was led by an angel to find some golden plates that were buried in a hill in Western New York. And that hill is in near the village of Calmyra. And this was an advertisement that appeared in two issues of the women's exponent of someone named Major Gilbert, who lives in Calmyra and is offering to sell authentic or genuine specimens of stone or wood collected from the site to members of the church in Utah for $0.25 or photographs for additional prices. The reason this ad is so interesting is that it contains several reversed shibbolets. That is to say that Major Gilbert in this advertisement refers to locations and objects using terms that the people in Calmyra would use, but would not be accepted amongst the general population of Mormons living in Utah at the time. The Golden Bible, for example, to refer to the Book of Mormon, or Mormon Hill, the Mormons would call it the Hill Camorra. And so this is a very interesting case of an advertiser not actually speaking the language of the audience that they're trying to advertise to. This is also interesting because at MHA this summer, one of the keynotes was given by the official historian of the Church of Jesus Christ of Latter-day Saints, LeGrand Curtis. And his entire topic was the church's preservation of this hill, near Calmyra, New York. And so I ran into him after another session and I pulled out my cell phone and was able to pull up this advertisement to show it to him. And we laughed about it. He thought that was fun. And when I turned to leave, he said, no, I want a copy. So he gave me his business card so I could send him a link to the database where he could find this, because he had, I mean, this is the official church historian. This is somebody who specializes in this specific topic. And yet he had never seen or heard of this advertisement before. All right, so other ideas. What other things could you do with advertisements? Oh, did I get the, oh, I got the Willard Marriott name wrong? Oh, I'll fix that. Super close. A super close. Slightly wrong. It's Jay Willard and I put Willard Jay. I'm so sorry. I should have known it's a Utah name versus initial first. Excuse my ignorance, but I don't know how far back like the official general conference practices were like that. But were there any connections that we could attack between advertisements that announcements made? Look at the top of the train table. Conference rates. Yeah, the Union Pacific Railroad offered specials in October and April for people to come to the general conference. Now, general conference today is generally two days. At the time, I believe it was five days long. It was, you know, they had meetings all week long. And so, yeah, no, there is here evidence of that. There are advertisements from clothing shops saying, hey, come see our fall collection while you're in town for general conference. And so, no, this is, yeah, I mean, 150 years ago, we were still doing that. So it's not just twice a year when you have to deal with traffic in downtown Salt Lake. That's a great question. And yes, we do have that all that went intact actually. I mean, let's conduct some original research right now. Let's search conference. So here's ladies who read the exponent. I will receive in time for the territorial fellow there at general conference, one of the largest assortments of books, stationaries, et cetera. So yeah, there's plenty. Here's, oh, that's interesting, for the conference trade stores West Side East Temple Street. I don't know what that means, but it looks like I don't know if that's referring to general conference or not. But yeah, there are several mentions of conference in these advertisements. So ladies attending conference will do well to give her a call. So visitors to conference. All right. So yeah, I'm also curious if you give these example of the one ad that's selling blocks of photographs and being that the person was an outsider, targeting the women in their current audience. I wonder if there were more advertisers, not local companies, local enterprises, but other outside companies, targeting for a reason. Yeah. And there were a few. The great majority of the advertisements were homegrown, but there are many of them. I haven't done that much work on them. We can browse entirely by vendor. So like there's the Mormon Hill excavation company right there. That's that one. The Oregon short line railroad would be another one that was from the outside. But by and large, these are these are locals. You know, desert tanning and manufacturing would be local. Bishop, Jasper Peterson, I believe you're just looking for Dunford. Dunford, I believe. Yeah. So you have A.D. Dunford and Dunford's. I think they're the same one, but they are shoes and slippers. So yeah, no, that's great. Any other questions or comments? I don't know if I have a question exactly, but I really liked hearing about contrasting sort of like the editorial position on on fashion with advertisements. And when I was kind of looking into women doctors of the time period, it was really interesting to me to see like the exponent coverage of some of these women and their ads. And then in some cases, editorials or things written by them. And so that that interplay just between how the exponent would sometimes feature these notable women, the women writing in it. And then advertisements, you know, mentioning them is just kind of an interesting area, I think. Yeah. And the Ellis and Maggie ship story is something that I had no clue about before I started this project, but just simply completing you know, their their advertisers. These are two women who had gone back east to get their medical education and then came back to practice in Utah. You know, that's almost like a movie kind of a mini series that I'd want to see on Netflix, right? Yeah, it was interesting. So this this database does not have the ability yet to connect, right? But, you know, we have redigitized an OCR the women's exponent. And it's currently last I checked, it was more than two thirds of the way hand corrected. Yeah. So so we're getting close to where we'll be able to, you know, it would be easy to link these these two things together. That would be amazing that put that in our mix. I was actually talking to the director of Eccles Health Science Library, and we were talking about the unique story of women doctors in Utah and how that's not really represented like on the walls of the medical school or library there, that it's just really a bunch of men. But we have such an interesting history here in Utah of, you know, it's like a lot of female doctors. I mean, at one point wasn't Deseret Hospital all women doctors? I'm getting nods. I hear you're getting beyond my room. Well, I work at the Super Plague Heritage Park when we have a desert hospital at Lake Ellis. Yeah. Okay. So I feel like I have heard that before. I mean, that's interesting history that I feel like got lost for a long time, but possibly could be like, we can in our, yeah, in our exhibit, we have a picture of the board. That's right. It's all women. So I'm thinking of ways that we could rectify that. Well, there are links that can be made. So for example, remember, I showed that Romania FAT had an office in the Constitution Building. Well, here is your own website. And there's a picture of the old Constitution Building. Number one hit on Google images, by the way, for old Constitution Building as well. We're so good. One funny thing though that came up though was, I mean, quite a ways down though, was one of the advertisements from my database. Oh, that's great. And it was, it was for, I can't remember. I think it was for a sewing machine with a wood cut illustration on it, because it was being sold out of the old Constitution Building as well. That's fine. And so, yeah. Okay. Well, great. We can move on to the next part of my presentation, which is the technology that has to go into this system. And this really hit me. Again, this was one of those moments of, oh, what have I done at MHA, because we presented in a digital humanity slash digital history session, including, you know, some people from, from this very group. And somebody asked the question, hey, that's really cool. What do I need to do to be able to do this? And I have been literally working neck deep in technology for 25 years, right? Being paid to do this, not counting my high school years of hacking around on an old IBM. And so I was, I listed out all of the technology that I used to build the system. And nothing earth shattering. These are not things that you need to have a doctoral degree in computer science to know how to use, but you do need to have experience of using them. But they are readily available. I use absolutely zero proprietary software to build this. It was all free software that came along. All right. So let's talk about the process of getting from the scams newspaper into the database. The first thing we have to do is we have to crop out all of the advertisers, 4,000 advertisers. The second thing we need to do is OCR, which is to take the image and trend and transfer into text, to a text file, then we need to group all of those advertisers by vendor and then by individual advertising. Then there's some hand correction to be done. And then there's a website to be built. The asterisk indicate steps where I wrote custom software. So every single step. Here's what I'll actually show you a live demonstration of this tool. This is, we initially called it Cropper, but that's actually the last name of one of our associate deans. So we changed it to Boxer, the Boxer website. This is just password protected so you can't visit it, but it just pulls up each paper. By the way, this is not the women's exponent. You might have noticed this is the next project that Elizabeth and I are working on, which is the anti-polygamy standard, which ought to show some interesting contrasts. But I would be, and this works really well if you have a tablet with a stylus, but I can, you know, I would pull up this, I would pull up a website and I would just click and click, and then I could click and click. And if I make a mistake and I get too much, I can just click on it and it goes away. And all of these clicks are being recorded in a database. So it's recording the coordinates of each box. And then when I'm done with this page, I can go to the next page. And you can see here are some bigger advertisements, which I would probably zoom out on to do. But just from a cursory glance of the anti-polygamy standard, they share zero advertisers with the women's exponent. Oh, two very, very different audiences, right? So that's the Boxer program, that just a simple website that we wrote to be able to crop out these 4,000 advertisements. Actually, it doesn't take long. Again, if you're doing something else, like, you know, I'm watching a football game, but in between plays, I can get four or five of these things marked off. And then it doesn't take long before you have all 4,000. Then there's a separate Python script that I wrote that will take that coordinate information for each picture and then make copies of those sections. Down here, as you can see, I'm actually adding some buffer. So it's enlarging the crop mark a little bit to make sure that we get what we're looking for. And then the next step is to do optical character recognition. What do you guys use here in the library? Abby or what are the OCR programs? Please, Abby. And we've also used Testoract as well. Yeah. So Abby, I find to give excellent results, but it's a lot more difficult to script. So Testoract, which is made by Google, is an OCR system that has a Python library. And what I do here, sorry, I'm going to give up some trade secrets here. But if you've ever done OCR on old documents, you know that the contrast level is super important. It's really important to get it just to the right point where you haven't lost enough detail that the letters are going to be recognized, but where there's not enough fuzziness that it's going to be confused. So what our system does is it actually runs each advertisement five times at different levels of adjustment. And then it counts the dictionary words greater than three letters long that it can find. And it assumes that the more dictionary words it finds, the clearer the transcription was. And so of course it takes time, but we can let it run overnight. And it auto-selects which adjustment level was optimal. So that's one of the benefits that come from being able to script rather than just having to use some sort of interface. So I was left with 4,000 advertisements and their text transcription. So it actually came to over 9,000 files when all of a sudden done. And now I get to go through each one and hand categorize it by vendor, by industry, and by advertising. Of course I didn't want to do that, right? My other option was this is I tried to protect his anonymity in some ways, but to call it old buddy from graduate school who works on the app to move my little panel down here. You guys can't see where we're going. He works at the National Bio-Forensic Analysis and Countermeasure Center. He's a geneticist. Just so you know, we can see it. It just blocks yours. It just blocks mine. Okay, so it's just up here. You guys can see it this way. All right. Yeah, you can see it on screen. Because there is a continuing relationship between genetic sequencing and text sequencing. We can use a lot of their methods in the humanities as we process large pieces of text. And so he gave me a couple of methods to use. I was actually a fascinating phone call to have and have just explained to him my situation and for him to say, oh yeah, that's exactly what we do. So what happened instead was I had misreported at one point that we used to card similarity index, but no, we used Levenstein distance. Levenstein distance is simply how many changes has to happen to a string of text to make it identical to another. Okay, so here we have four sentences. And here we have a table that shows how many changes would have to happen to make text A become text B and so forth. Now, if you attended David's Gephygraph seminar a few weeks ago, you know that we can plot these distances. And so he, oh, by the way, it is computationally intensive, because instead of having, oops, here we have four texts by four texts. So we have 16 crosses. I have over 4,000. So that's 16 million cells in this table. Right. So the computer goes crazy while we're doing it. But this is a graph that shows all 4,000 advertisements plotted out based on their distance from one another in terms of Levenstein. And so, you know, these are probably, if not the same vendor, they're the same industry because they're using very similar text in their advertisements. And then of course, if two texts were identical, they would appear directly on top of each other. Well, you also know from David's presentation that the computer can detect communities. That is to say that if some of the nodes on this graph are highly interconnected and less connected with the nodes around them, that they might be a community. And specifically, I use Louvain Community Detection, which if you go to Wikipedia, this is the algorithm board, which I simplify to this. Okay. Because there's a Python library that will do Community Detection Force. And when I ran that on it, it separated the 4,000 advertisements into 172 groups, plus some that I labeled orphan. And I'll show you what those look like. And it worked really well. That is to say that the computer was able to group all of these advertisements together, especially when they're word-for-word the same. Formatting doesn't matter because we're dealing with the text, not with the font. And so, all of the computer recognized that all of these belonged together, even though they changed the format of the advertisement. Now, it's not perfect. This is, here we have Ellis ship, Maggie's ship, and I believe there's a Romania Pratt in here somewhere. And the doctors ship down here. But the computer found enough similarity in these that it put them all together in one group, representing more an industry than an individual. I would still rather go through 170 folders with that have been pregrouped and just correct that, then have to do a complete categorization using 4,000 images and text files. But there's a cool part about that Levenstein algorithm that we use, which is that community detection is relative. That is to say, you might, if we did, everybody in this group, if we did your Facebook friends and ran community detection on it, we would see who are David's friends and who are Comstock's friends and so forth. But then we could take David's group and run community detection on it as well. And we would find who are his cousins who are his immediate family and so forth. And so I could rerun that community detection within the group and get a different result. So this is on that midwife group. And it found 15 subgroups inside of it, which was a lot more accurate. Here you can see all Maggie's ship. And then, not perfect though, but here's why. So on the left, you have a class in obstetrics by Dr. Maggie C. Ship. On the right, you have one by Dr. Romania B. Pratt. But look at the text, will commence a class in obstetrics on versus will commence another class on obstetrics about. So the computer thought, these are similar enough compared to how different the other ads are to put them together. So it's not perfect, but it greatly accelerated the rate at which we could classify these advertisements. All right. In terms of the orphaned advertisements, if you look at them, they're largely damaged or highly illustrated advertisements. And so the computer had a hard time finding text and therefore couldn't attach them to any group. So that's the advertisement that pops up when you search the old constitutionally. Okay. So hand correction. This was another fun part. Like many areas of campus, we employ, we employ a number of students in our office who oftentimes are waiting for people to help. And so we will sometimes give them tasks, whether it's transcription or review of something to do in their downtime. And so this was the interface we created for them for that. And the idea is here are all of the examples of this advertisement. And we would like you to click on what you see as being the exemplary version of it. And then down below, there's the transcription provided by Tesseract. They corrected that. And then we have a few fields down here which we haven't done anything with yet. But they put the names of people that are mentioned, the address if there is one, direction, excuse me, if there are any, and a telephone number. So the students were able to go through all 300 advertisements that way and do some hand error correction. Okay. And then finally building the website. And I had, you know, I could use any of the free website options out there. And I looked at all of them and I said, nah, let's roll our own. I was feeling a little uppity at the time. I think I was depressed because we had had some plugin on WordPress break. And I was tired of depending on other people to do it. But something that amazed me because it's been so long since I had written a website for the ground up, despite the fact that I teach classes on that, is just how simple the code gets when you write it yourself. Because it's for a very narrow case, it's not generalized out to anybody else that might want to use the same system. So this is the entire code here for that index page. You know, it's 38 lines. Now I am including other pieces of code remotely. But you know, it was pretty easy. It wasn't hard at all to do. I mean, for somebody who's experienced. And better is that I have the feeling that this code is going to be stable for longer than the other generalized systems would be. Because I'm using very few functions and they're all locked down types. You know, it's kind of like the difference between a canoe versus the Titanic. So, you know, that's how I built the website. Questions, comments, ideas, criticisms. My comment is, it's very cool. And thanks so much for showing us your work. And I'm wondering, like, where, where you see this going next. Yeah, so we have, so there is a feedback section. There's a form on here for people to leave comments, whether they find errors, such as the, it's Jay Willard and not Willard Jay. And so that I can go in and correct that, or suggestions, or just general questions. Nobody has used that format. I haven't gotten a single comment for it. And so that's, that's a little frustrating, but at the same time, and I'll show you in a minute, that this has made waves because it is a cool project, right? And I think it was worth doing. But for next steps, we'd like to add the anti-plugmas standard there. You know, and see, I can use the same interface. All I have to do is add a field for, you know, which, which newspaper this came out of. And, and we'd be able to keep them separate, to do comparisons and so forth. It'd be really interesting to do that. Now, I am not a historian of Utah by trade. And so I'm not the person who would ultimately use this for that kind of research. And that's probably it's biggest weakness. So, yeah, I appreciate that. Thank you for saying this. Her area of research is on the women's exponent and anti-plugamy standard. You know, I think Elizabeth has, I have not directly though. And so that would be interesting. I've been in them, I think one meeting with her, but it didn't come up. She'd probably be a great collaborator because this was like a very specific area of interest for her. That's a good point. I do have to say that there are several faculty in our religious history department who I've done the whole OCR thing. I've spun off to a staff member who's an awesome developer named Jesse. And he has been working to, to OCR a huge number of publications from the community of Christ, which was the reorganized church at Jesus Christ of Latter-day Saints. They had a magazine or they have a magazine that they published consistently for over 150 years. And so we, and it's publicly available, but in image format. So we have that. The Young Women's Journal, which was another publication here in Salt Lake, we've done. And so we do have members of the history of the religious history department at BYU that have been taking advantage of similar things like this. But you're right. I haven't quite approached them yet to say, Hey, what do you think about these advertisements? So that would be interesting. They thought about wrapping up the software and the package. It could, it could be, but that wasn't my goal when I created it. There is a lot of overhead that goes into, you know, you know, putting it up somewhere so that somebody else could load their data into it. I think that the tools that I use to create the data set would be much more amenable to that than the website itself. So for example, that, that, that boxer program where, you know, it's an online system for cropping out things on, on from a collection of images, that might actually be a prime candidate to go on to get help or something like that so that other people can use it. I'm not done. Not done yet. So I put up this list of what I do. And this is, this is something that's been weighing on my mind pretty heavily since 2015 and you'll see why. But this characterization of digital humanities work, it gets attention. Bigger tends to be better, right? Your corpora are measured in billions of words or number of documents that you have in your archive. And the impact is often measured by accounting users or citations or something like that. And we still have competitive grant funding, which is zero sum. And we still have this territorial owning of space, right? I mean, this is a great accomplishment to have your own space. And that just screams to me, I'm sorry, but toxic masculinity. And, you know, thankfully, the digital humanities from its inception in the early 2000s has always been open and collaborative. But it's still, we're still fighting against some of those, you know, more traditional or, you know, ancient regime ways of doing things. So in 2015, I was at DH in Australia. And I have been searching high and low for the first thing he said this, but I'm afraid that they didn't actually put it in their paper, which is why I can't find it. But it was a woman presenter who talked about big data and DH and men comparing the size of their tools once again. Okay. And then this is a great book, if you don't have it, look it up. It's Liz Lash's edited volume on intersectional feminism in the digital humanities. And it provides some really interesting arguments against a lot of the ways that traditional academic procedures bleed into the digital humanities. This is from, this is a quote of Martha Nielsen Smith, it's in one of the chapters. These matters of objective and hard science provided an oasis for folks who did not want to clutter sharp discipline methodological philosophy with considerations of gender, race, and class determined facts of life. Or Rupert Boyd and Howe who explored what queer thinking could bring to the digital humanities. And then this last one is the one that most resonates with me in this project is challenges of trying to do anti-racist, coalitional feminist work in the context of non feminist institutional structure. That is to say that many of the criticisms that I would have of this project that I did myself is it's small, it's parochial, it's got limited applicability, no one's ever going to look at it. But at the same time, those are all criticisms that come from this sexist model that we have in the institution. So if we look at these collaborations that I've done, none of these projects that I listed were independent. They were all collaborative in some way or another. The stars are the collaborative ones, so they're all starred. And then I put a little Venus symbol next to the to the women collaborators that I work with and a tall S next to ones that were student schedules. And this is probably what I'm most proud about is that I've been able to reach out across campus and help so many people with their reasons. And that makes me wonder if we don't need different criteria for impact in the digital humanities. On the left is the multi-media dictionary of Ketua ideophones, which I worked on with Dr. Janice Knuffels from our linguistics department and her students. If this site goes down, it'll be two weeks before I hear about it. Okay, because there are probably only seven people in the world who have to be using it. But when I just two years after we started that project, when I asked Dr. Knuffels for an update on what she's doing for it, she listed four articles slash book chapters that are either published forthcoming or in progress. And then I think seven different presentations that she'd given in the matter of two or three years. Because once her data could be organized in this manner, it became super impactful to her. So do we need a different measure for impact? The one on the right is the 16th century French religious pamphlets project. Yes, again, you can probably fit everyone who's interested in that in a small room. But the professor I helped out with this based his tenure goes to AI. It was hugely impacted. Similarly, this is a project that I did last spring with the University of Utah Masters of Public Health students on basically doing text analysis on 10,000 articles and news broadcasts referring to masks and mask mandates during the pandemic. I should have looked at the professor's name, but the professor was so appreciative of this work that she suggested that the student Byron Montgomery and I turn this into a paper and submit it to the American Journal of Public Health. And again, we haven't done that yet because Byron is now a certified practitioner in public health and he's waiting until next year so he can get professional development credit for writing a paper. Again, hugely impactful to him, but unless I go to some public health conference, I'm never going to be presenting on that myself. Similarly, you have this project. The BYU News picked it up around Pioneer Day. They thought this was a great project to highlight during Pioneer Day. So there's Elizabeth and myself, and they ran a story on it, and then abc4.com picked it up and wrote a piece on it. Why? Well, because it's interesting to people around here. Even if it's not a ton of use, it's not really terribly useful to historians. It is interesting to people around here and for a moment cast some of their minds back to the Pioneer heritage that we have here and some of the struggles that women in particular faced in early Salt Lake civilization, our societies. So that's what I've got. Thank you so much for your time and your attention. I appreciate it. Jeremy, I think from a library science perspective too, I see so much of what you're doing and it feels like collections is data work. You're taking collections that we already have and figuring out how to make data sets that people can play with computationally. So I think to different audiences, your work is going to mean different things, but both with your mass mandate work and historical newspapers, you're turning library collections into data sets. Yeah, and what is the chicken MEA in this case, right? I mean, when you're in library, how do you determine, oh gee, it's worthwhile to go through and tag all the metadata on this on this collection because somebody could use it or is it, hey, this is low hanging fruit. So let's tag that and what's the process for that? I don't know if we have a real structured process because so far when we've been sort of doing similar work developing, you know, small or large data sets, we've looked at things that appear to be interesting to us, you know, and done the extra work and release them. And in some cases, folks have done some things with that. And in some cases, I don't know, you know, so not yet, not yet. But I think it's, I think it's worth going through this kind of process because it really, in my case, you know, I start thinking about digitized collections in different ways and different potential uses of that, other than just putting something online in a digital repository and calling it done, you know, so that's kind of a non answer. Any questions on line yet? Really fascinating work. Okay. Oh, I'm not going to deny that. This was one of those projects where I never actually had to say, oh, I got to work on that project again. Right. And I think part of it is because it's not an area of content matter specialty for myself. And so I'm coming up with questions that I'm asking Elizabeth, these questions, like, well, I just found this, what's going on here? Or, you know, Googling Romania Pratt, I said, wow, that was a fascinating woman. You know, and so the fact that I was learning something every step of the way, just, you know, it kept me going. I think a lot of us on the women's extended project felt that way. It was really fascinating content to dig into. And I just kept thinking, why don't we know more of this stuff? If we know a lot of the male leaders in the early Mormon church or hear their names or lots of things are named after them, but I had never heard of Romania Pratt or Ellis Schiff or any of these. Okay, so I'm going to reveal something to you, but Romania Pratt, I had never heard of, but do you know where I had heard M.L.D. Wells and Schiff and Richards and so at BYU, they torn them down, but there used to be a set of dormitories that were the apartment style dormitories. And historically, that had been an exclusively women's dorm area. And so all of the buildings were named after historical women in ad, but they've all been torn down and replaced with numbered buildings instead. And so, but yeah, you're right. We don't know much about them. I mean, but the fact that the only thing that I had known about them was that this building where, you know, a girl I dated when I was an undergraduate lived was, you know, that was the name of her hall. Yeah, that's, you know, and so then the next question is, what can we do about it? Right, you know, I would think with this deseret hospital thing, do they have a place where they have pictures of all of the boards? Right, because if not, hey, let's, you know, and even if the boards at some point transitioned into male, right, we can go back and show that there was this predominantly female board or an entirely female board in the beginning, you know, to draw attention to the the feminine origins of the institution. And yeah, not just having the image of having it be super visible in the hospital, the medical school or library, because I'm sure it exists in like the medical history room, but it's not being. And although these women offered mid-wide services and training for midwives, they were not nurses, they were not midwives, they were physicians, they were full-fledged licensed MDs. And, you know, a lot of times when we see these pictures of women from, you know, the late 19th century in a medical situation, you know, our bias tends to project them into one of those roles. So, I really, I can't get enough of those stories you found through this. Could you text in your research, because the women's experts continue at some point, was there a shift in advocates that led to the choice of the paper or made a much more formal decision? I don't have any evidence on what led to that. It was replaced by the Relief Society magazine. I just, there was such a long period towards the end of the run, I mean, but it was decades long that there was very little advertisements. Yeah, I mean, the last decade, we're talking one or two maybe three advertisements per issue. And so I'm thinking at that point, they might have been, they were either supported by subscription fees, by, you know, direct support from the churches and institution, or they had started running the deficit. You know, some, some, there's some explanation like that, but I don't have any other evidence to explain that, but there is this obvious diminishing there, they're advertising revenues in. Was there like one holdout that kept us for the past 10 years? Yeah, I mean, one of the, you know, because we can go to Browse, and we can go to, thank you for working. Yeah, Alice Ship and R.K. Thomas, who was a, you know, clothing sales, but yeah, ran through many, many, many of of the issues. But yeah, you can see that they only had two advertisers in that last year, 1913, Alice Ship, R.K. Thomas. Oh, and then this, this map for Book of Mormon study, which is funny because I said, this is a hot topic at the debate on Twitter right now too. So something's never changed. But yeah, you can see they only had a, you know, just, you know, you could count their advertisers on one hand through the last several years of their, of their run. So, yeah, that's a topic to investigate. Why, what happened? I have something I love about your projects. It's not that they always are answering the questions, but they're pointing to questions that could be investigated. Yeah, and Dr. Cole, what's Dr. Cole's? The Poelmage Project. Hey Cole, yeah. Her and I had an argument in 2017 at DH in Montreal. We were on the same side of the argument. We were arguing with a young European graduate student who had attended my session where I made this point. And it's not original to me, Jeffrey Rockwell and St. Clair, who created the Boyant Tools. They see digital humanities as very much a playground. But just like kids learn on the playground, we're learning from playing with these tools. And the questions that they develop in our minds that drive us deeper into the text are what's really valuable. And so the argument that this, you know, in Europe, the digital humanities is seeing much more from a positive lens. So he did not take well to this idea that, you know, the real value is in generating the questions that we can then, you know, go back to the text to try to answer. But yeah, that's exactly what this kind of project does. So you explore these things and the questions start to pop up in your mind. Well, do we already think from the zoomer? Let's take Jeremy once again. Thank you guys. And yeah, Boyant Tools does rock. You should do another session with the sun going off. It's a lot of fun to do that. Yeah. All right. Thank you.