 for many historians, architects, and students and teachers for many years to come. On what we have shown you today is an interactive map. We have made sure that the data we have curated, and here I'm using the term curated as a technical term, for preserving and properly documenting data. We have curated the data so that it can be downloaded for research purposes to produce histories and analyses in many other formats. But today, we are very excited to show you our map, which is only the beginning of much more work to come. Sorry. First, a lot of thank yous. The work would not have been possible without the support of the Robert David Lyon Gardner Foundation. Thank you. Thanks also to Amal Andras, Dean of GSAP, Adam Costow, the Chair of History, and the Institute for Social and Economic Research Policy, ISERP, for administratively supporting our work. Thank you as well to the Empirical Reasoning Center at Barnard, Research and Data Services at Columbia Libraries, GSAP IT, and CUIT. All of you have helped in various social and technical ways to support our work. A special thank you as well to the Andrew W. Mellon Foundation, which supported the launch of the Center for Spatial Research, a hub for urban research at Columbia that links design, architecture, urbanism, the humanities, and data science. It sponsors research and curricular activities around built technologies of mapping, data visualization, data collection, and data analysis. We focus on data literacy, on interrogating data, and working to open up new areas of research and inquiry with design tools to help scholars, students, as well as our collaborators and audiences to understand cities worldwide, past, present, and future. This project we are launching today could not be a better example of what we can do. During the first years of establishing CSR, we ran summer boot camps for Columbia faculty and PhD students to learn spatial analysis and mapping tools for their research. Rebecca Cobran, Russell and Bettina Knapp, an associate professor of American Jewish history, was first in our cohort. So in May 9, lung professor of Asian American studies and professor of history and co-director of the Center for Study of Race and Ethnicity approached Rebecca to work on a map of New York City. Rebecca brought CSR into the mix and encouraged me to take the boot camp the following summer. Then the work of writing our proposal began. What followed when we received funding is something that happened so easily at universities. We assembled a team of experts on the topic. Gerga Boych, associate professor of history and urban studies. Is a historian at Barnard who specializes in historical GIS. Leah Meisterlin, a sister professor of urban planning at GSAP who specializes in GIS methodology. Sarah Broly, who is associate director here at CSR was key at the beginning phases of the project in leading boot camps to encourage the kinds of collaboration that this project exemplifies. Her leadership in the center and approach to methods of spatial research set the tone for our collaboration. Then we ran a search to hire a postdoc, Wright Kennedy and a research associate in architecture, Dan Miller who you will meet during this launch and be witness to the incredibly detailed and rigorous work they have done in assembling the map. They led a team of 40 students from across Columbia, architecture, urban studies, data science, computational social science and beyond. Thank you to each and every student involved. So here we are showing you the hours and hours and hours of work it took to geolocate the historical census onto a map of New York City as it evolved from 1850 to 1880 to 1910. You will hear later that the census is not a perfect record of population. So there are uncertainties in it as with most data. Once we were far enough along with the work we engaged with Stamen Design who assembled this work into the gorgeous interactive map you see here today. Thanks to the Stamen team. We are excited to announce as well that our work will not stop here. The Gardner Foundation has generously renewed the grant so that three years from now we will have mapped the whole city as well as Nassau and Suffolk counties in Long Island for 1820, 1850, 1880, 1910 and 1940. This also means that our team will keep gathering right here in this room for bi-weekly meetings to continue this extraordinary and fun collaboration across disciplines and fields of history, architecture, urbanism and data science. More students will join us too. We hope that as the project is made public that others at the university and well beyond will use the map as a resource and suggest or create more case studies to add to the map. Please do contact us. We are serious about expanding our work. And now I'm going to turn the presentation to May Nye who will give you a brief overview of the map itself. May, come launch the map. Ready to launch? Over the three year course of this grant our team created web-based interactive maps of Manhattan and Brooklyn to show changes in demography that is race, nationality, gender, marital status and occupation and land use from 1850 to 1910. The map breaks new ground by locating each person counted in the census at their home address sometimes before the street grid was even established. New York changed tremendously during this period especially around the turn of the 20th century for two major reasons. First the influx of immigrants and second the inclusion of Brooklyn the fourth largest city in the country in 1890 into New York. Our project visualizes the magnitude of these changes that took place during this time. To do this we used historical maps and city directories and even traced the census takers steps. We employed dozens of student researchers and used data from the decennial census, geolocation tools and geographic information systems or GIS. You all know GIS it powers Google Maps on your phone. GIS is a revolutionary framework that allows for the visualization of data population characteristics, the built environment, industry, et cetera in spatial form that is maps. It allows researchers to discover patterns and trends that will be impossible to grasp through statistical tabulation alone. Much more than cartography, GIS allows for the investigation, manipulation and analysis of multiple layers of data and hence the understanding of complex spatial relationships. Today massive amounts of data are available to facilitate urban studies but their effective utilization has been hampered by two factors. First, urban historians and other scholars in the humanistic social sciences have traditionally been the least methodologically equipped to take advantage of the explosion of data that are now available. And second, while searchable micro data of all types for our time are abundant and the use in GIS applications is now common, historical micro databases are only slowly being made available, especially for any period before 1970. Our project is notable for its ambition to map systematically and sequentially the big picture of all of New York and the micro level, that is individual residents and block level analysis. Previously only ward level census data could be visualized. Our aim to map at the household level pose special challenges because addresses are not usually available in searchable databases before 1900 and sometimes do not even exist in the modern sense. Geocoding at the individual level provides data at the smallest possible level which can be aggregated at any analytically meaningful spatial unit allowing for the most robust research and analysis. Our project was inspired in part by the maps created in the 1890s by the legendary whole house settlement in Chicago led by Jane Adams. The whole house residents which included women sociologists went door to door and interviewed people in the near west side of Chicago a working class immigrant neighborhood. They created maps showing the distribution of nationalities and wages in the area. We did not go door to door of course but we relied on the work of census enumerators who did at three points in time. These maps are both a research tool and a teaching tool. The project offers a valuable opportunity for Columbia faculty and graduate and undergraduate students to learn and enhance their own skills in digital methods. Our hope is that these tools can be used to teach historians and their students not only the historical content of the material but also new techniques in digital humanities. We're very excited to show you more. I want to now introduce you to our team members the right Kennedy and Gurgle Beish who will discuss the potential of these maps for research and teaching. Then Leah Meisterlin and Dan Miller will present the backstory of how we created the maps. And finally, Rebecca Cobrin will host our Q&A sessions. Please send any questions you have to the Q&A function on your webinar at the bottom toolbar at any point during the program. Thank you. And now I bring you right Kennedy. Thank you, Mae. We designed mapping historical New York to make the city's history more accessible to scholars, students and the broader public. We designed mapping historical New York to make the city's history more accessible to scholars, students and broader public alike. I'm excited to share with you some of the ways that our Atlas will facilitate scholarship concurrently, providing both the digital tools and new perspectives on traditional research questions. Afterwards, Professor Beish will illustrate the power of the Atlas as a teaching tool. We provided three avenues to engage, we provided three avenues to engage with historical data set. The case studies highlight map options and functionality while illustrating research questions and narratives. You can produce with the digital Atlas. To locate a case study, click on view a case study and then you'll see the stories on the side. Dan and Leah will talk about this case study showcase later. A second more flexible way to use the Atlas is the explore tab, which will be the focus of my talk today. Here you can customize the map views and layers, examining a range of topics, such as race, gender, age and birthplace. Third, we are finalizing a data download page for advanced users. You'll be able to download the summary level tables and geographies along with the location data for each person. Today, we'll use the explore tab to create custom map views to locate Italian neighborhoods in 1910 and we'll examine the boundaries between these neighborhoods and other immigrant communities. Let's first look at the community boundaries. Each dark blue dot represents a resident in 1910 who was born in Italy. Zooming to the lower east side reveals the proximity of Italian residents, again the dark blue dots and the Russian Jewish residents who were represented by the yellow dots. If we zoom even closer to the street level, we can see areas of transition between the communities for Italians and Russian Jews lived on the same block. Even in these heterogeneous areas, however, the map shows that Italians and Russian Jews tended to live in separate buildings. By switching over to a summary geography, researchers can visualize these patterns in more quantitative ways as well. Summary geographies can help us locate the Italian neighborhoods of 1910. The map now shows the percentage of Italians at five ward. The spatial patterns can look very different depending on the geographic units. We've included multiple options in the address. If we switch from wards to census enumeration districts or to blocks, we see a very different pattern. The map now reveals an Italian neighborhood in east Harlem which wasn't visible in the ward map. This built-in flexibility to move across summary units gives researchers more tools to reveal and investigate the spatial patterns at the city's residents. Likewise, researchers can quickly spot areas of interest for other topics, such as race, using summary geography, which we see here, or at the dot density view. In this view, purple dots represent black residents, orange dots represent people of mixed race, and blue dots represent white residents. The map makes it easy to see a community of black US-born residents and mixed-race Caribbean-born residents living around 134th Street. Just a few years after the 1910 census, this small community boomed with the Great Migration and black Harlem emerged in the 1920s and 30s. These are just a few examples of the ways in which you can use the Digital Atlas to ask research questions, investigate historical topics, and contextualize historical interpretations. If this isn't enough, we will offer data packages that you'll be able to download and use with desktop software. This will include the summary level data and geographies in addition to the home locations for each of the 6.5 million residents. The person level data consists of latitude and longitude along with a unique identifier that can be used to join any additional variables from the full-count census hosted by the Minnesota Population Center and Infos. This is especially exciting because you'll be able to intersect multiple variables of interest and still map these by the residential locations. Moreover, you can bring these various layers together and bring in your own layers as well. So you can bring in schools or redlining maps for comparison and analysis with the census records. For example, you might download the individual location datasets from our Digital Atlas, join it with the full variables from IPUMS, map the historical locations of grade schools, and use spatial analysis to measure the average distance of each black family to grade schools as one metric of the educational accessibility in 1910. We hope that our project makes the historical censuses and geographies both more accessible while also lowering the barriers to entry for studying the city's history, our shared history. We can't wait to see what you discover with the Digital Atlas and the historical datasets. Now I'll hand it over to Gerga. I think the story of New York has equally great potential for each. You've been thinking about audiences in the sense that circles higher education or more broadly grade education, especially at the high school level and even more broadly the general public from museums and local historical societies to individual users. To begin with higher education, mapping historical New York contributes to several historical fields, including urban history, immigration history, race, New York City history and spatial history. It offers different levels of possible integration into the classroom. First, it can be a valuable teaching tool in lectures and seminars to explore and visualize spatial patterns and open up discussions. Second, and more intensive, spatial history modules and assignments can be built around the website for individual or group-based projects. We have an ongoing partnership with the Barnard Empirical Reasoning Center which supports such modules in a range of courses. In this case, one such module could have students working groups on case studies like the ones already available from the website. We hope the case study function will eventually become a compendium of spatial explorations leveraging the website. Third, and even more intensive, an entire course can be designed around mapping historical New York. Such an approach is most appropriate for a spatial history class. It likely involves teaching both historical content, for example, immigration history and historical GIS in a seminar plus lab or lecture plus lab format. In fact, Wright and I are offering such a course this spring. In my experience, the biggest challenge in teaching spatial histories to help students locate or create spatial historical data for their projects. This can be extremely time-consuming. It can also present a significant barrier for faculty unfamiliar with historical GIS. Mapping historical New York provides a resource to overcome this barrier, whether one relies entirely on the website or makes use of its curated data. Let me offer one example looking at residential patterns by gender. The historical census identifies all residents in a binary division of male and female. Looking at the 1910 maps at the block level showing female New Yorkers on the right side and male New Yorkers on the left side of the screen, you may notice some patterns. Most evident in Upper Manhattan, women represented the majority of the population, whereas along the shore, we see the majority of men. How do we explain this? The website offers some visualization options to interpret these findings. For example, we can select a dot density map that shows female New Yorkers in green and male New Yorkers in pink and zoom on individual buildings, some of which are inhabited entirely by women or men suggesting single sex boarding houses. We can then return to block level summary geographies and explore select occupations like domestic servants for women on the right side and laborers for men on the left side of the screen. The high concentration of domestics in Upper Manhattan corresponds with majority female blocks. One should note that these were middle and upper middle class neighborhoods and domestics typically resided on the premises of their employers. Further inquiries may look at the birthplace category to consider specific immigrant groups or explore age. For a full-fledged analysis, students in a spatial history seminar can use the downloadable data and dive in more deeply. They can intersect relevant variables or isolate certain groups such as young Irish women domestic servants. With GIS competence from a lab, they can conduct additional spatial analysis. Mapping historical New York is also well suited for grade education, especially at the high school level. In this context, its value extends beyond learning about specific areas of history to provide a tool to teach spatial literacy. At the high school setting, the most effective strategy is probably to develop thematic lesson plans and learning modules based on the website. Such lesson plans can focus on integration, race or gender. Students can study citywide patterns or focus on specific neighborhoods, blocks and streets. We plan to workshop with high school teachers to design such lesson plans and make some available as examples. Most broadly, we hope mapping historical New York reaches a wide range of audiences. We hope to see our website being integrated into exhibits, walking tours and research and teaching project based at local historical societies. Likewise, we seek to engage individual users. For that reason, we created a user-friendly interface for people to make their own individual discoveries. This can start with questions as personal as who lived in the building where I live today? What were the other buildings next to mine? Are they still there or how did my neighborhood change over time? In fact, Skirmorhorn Hall on Columbia University's campus, the building shown on this 1911 map is where we're all located right now as we launch the website. Ultimately, we hope mapping historical New York can become a useful resource to foster greater connection to New York City and its history. And I will give the floor to Lea Meisterlin and Dan Miller who will present on the backstory of how the maps were made. So sadly, we cannot cover the full story of how the map was made. Suffice it to say though that the digital Atlas and the data it contains was centuries and three years in the making. We will cover a few highlights with a few examples but please know that we'll be summarizing and just barely scratching the surface. Of course, we'll be happy to answer questions and full documentation will be available on the website. And while our team is rather large, we also wanna take a moment to acknowledge that the full interdisciplinary team extends well beyond the seven of us here. With an advisory committee that reaches across the university and the city and research assistants from fields including urban planning, history, and computer science. Seeing digital Atlas was centuries and three years in the making is a short-hand way of noting the accumulation of information represented in the map as well as the labor. Introducing Dan Miller. Hi. I was just talking about how seeing that the map was centuries and three years in the making is a short-hand way of noting the accumulation of information represented here in our website in the map itself but also the labor required to produce it. So starting with the census from the politics of its categories to census takers canvassing through decades of cumulative mappings of the city to the present day worth of scanning, digitizing, and reconciling this material. This list on the slide includes much of the source material required for the interactive Atlas. And it's important to note that as a result the map reflects at its best who was recorded in the census which is not the same as who lived in New York. We'll discuss our approach in terms of a double-sided process divided by the types of data we need to locate addresses within the historical city. On one side, we obviously need clean and machine readable census data. On the other side, we need historically accurate information about the city to locate those census records. After all, our street system and building numbers have changed dramatically since 1850. Think of it this way. We need to find or create reliable reference data about a city we cannot visit without a time machine and we are working with census records so we can't exactly verify either. This two-sided process involves research activities that we can organize into five broad categories. Those are digitization, standardization, information recovery, or filling in the gaps, refinement, and finally, web photography. We think of each of these categories as spaces where the methods of history, GIS, design, and data science meet each other in different ways. So for example, under digitization we should start by acknowledging that we are building on the Herculean efforts of Iphons and the New York Public Library. Iphons has digitized the historic census manuscript records, which you can see on the left, and the NYPL map worker project included dozens of scanned and geo-referenced historic map sheets. Both of these served as important starting points for our work, including drawing or digitizing the city's historic streets, as you can see on the right. Unfailingly, of course, with so much material digitized, we will find inconsistencies that need to be standardized before we can move forward. A couple of examples include standardizing names, filling census records after text recognition attempts to transcribe different census papers handwriting that's on the left, as well as how to handle the variation in street names, such as the case you see on the right of Elizabeth Street, which is also known as Beard Street. In the case of these streets, we work between these two categories of census and reference, and developed an algorithm that compares the street names founds across all those listed in our digitized census records. But even with digitized and standardized records and reference material, we still adapt to fill and information that needed to be recovered. And here is where comparing many different historical sources becomes not only useful, but necessary. For example, on the left, where addresses were missing in the census records, we used what is called a fuzzy matching algorithm to cross-reference these records against city directory listings. This allowed us to fill in and then map more of those records. And rather than relying on one map for each year, we referenced several including plans for the development yet to come in the decades between our census years. Making these comparisons between base maps is also possible through the Atlas that's shown in the image on the right. This helped account for more missing information and added to our growing list of layers in the Atlas by detailing the streetscape, creating historical shorelines and so on. And eventually, as we fill in these gaps, we start to feel pretty good about the data sets on both sides of our process. We felt confident in the quality of our census records, as well as the reference material we used to map those historical addresses. And then you plot those addresses. And inevitably, there are always addresses that don't find a match on the map. And so we reflect upon why that is. Refinement is the long part of the process. Going back to our earlier stages to refine our material and revisiting both sides of this two-sided process. For this project, refinement meant that digitization, standardization, and information recovery didn't just happen once, but we're ongoing at two weeks. So a couple more quick examples. First, for the census records, again, on the left, we realized that including enumeration dates within our digitized records could help us order and then model the paths of each census taker, as they build in their questionnaire pages moving from one house to the next. This extra digitizing, in turn, allowed us to start plotting some of the records with missing addresses between located points. And for our reference material, we extended the work of the New York Public Library last team who developed a method for applying text recognition within accounting directories. As we move forward with the Brooklyn Center and directories, we built upon their method to account for the particularities of the type setting. For example, in the images on the right, you might notice that the printer of the city director develops an alternative punctuation conventions for whenever they ran out of commas in a particularly large printer. Finally, since we're describing how the map was made, we have to talk about making the map. What the web map shows is what the census recorded and where possible, but we were able to recover or add through some careful methodological innovation. Again, we can't walk through each stage of the cartographic process. So we'll go through just a couple of examples where research and map design come together to offer a little more insight. First, especially in the earlier years and more rural parts of the city, the census records don't include as many mapable locations. In other words, we have reported residents in rural regions often without streets or without street addresses. Such as Southern Kings County, as pictured here. As a result, the map offers different ways of visualizing the census layers, showing mapable addresses where these buildings were reported, as well as full counts records organized by different areas. Here, for instance, we can see the percentage of local residents born in Ireland in 1980, although the map buildings where they live are limited to the more urban northern part of the city. This is a great example of what we need by noting that the map reflects what the census recorded rather than who lived in the city. Another example of where the cartography intersects with the research process starts with where we are able to confidently plot census records at specific buildings. And here, Stamen's contribution as a design partner has been invaluable. Pulling from urban cartographic history, including references from the whole house maps, the Atlas features a building level dot density map of each census category, which we've all shown a little bit of. Each person is represented by one dot and those dots centered around their respective building address. To our knowledge, this is an unprecedentedly high resolution census visualization, especially in covering such a wide area. This allows us to ask questions about density, diversity, segregation, and social stratification, not only across neighborhoods or across the city, but block by block and even within buildings. And with that, we'll hand it off to Rebecca, who will wrap up and moderate any questions and discussion. Thank you, Leah and Dan. And I wanna just thank everyone who's here. We invite you to share your questions. I have used GIS mapping and earlier versions of this map in my immigrant New York class and I have seen up close. This is a very powerful tool to force students to ask new questions. This website's interactive features not only allow us all to map and visualize residential geographies of both New Yorkers and Brooklynites by their race, ethnicity, gender, place of birth, and occupation across the late 19th and early 20th century. But it also allows each of us to zoom in and out to view the whole city, neighborhoods, and individual buildings. By enabling us all to see simultaneously both the city-wide level changes as well as individual level changes on specific blocks, this map will enable you to see spatial patterns and trends that would be impossible to grasp by looking just at census material or historical maps. We see this map as a conversation that will grow with your use of it and your questions and your insights. And now I wanna start that conversation and I ask you to continue to send me your questions and what we're not able to answer in this hour, we will get back to you. But before I go to the questions, I wanna say the link to the website will be on the CSR, that's the Center for Spatial Research website, and Columbia's History Department website after we end this launch. But I'm now gonna start with Eric Chow, the first question you get answered, who asked us a question about what is the maps that we're using for this project? API maps, Google maps? I think it's an excellent question to start off this whole discussion and write. Will you take over? Yeah, so one of the really interesting things about this map is that with the help of all the research assistants and all the students that contributed to this project in addition to the expertise of the principal investigators, we were able to reconstruct, they reconstruct a model of that historical world. So we're not using Google or another map API, we actually built out historical coastlines, historical park layers, everything that's very temporally sensitive. So it matches up with 1850, the coastline to the best of what we could get off the historical maps match up in 1850. So all those base layers that you see in the website are custom built by our team to reconstruct these historical places in this historical world. Thank you. So Monica asked a question about fuzzy mapping, and I would like Dan to answer this because I think this is in many ways one of the greatest interventions that this project brought about of how to think about how we use historical sources to map across time and space, Dan. So the method that Monica is asking about was developed out of the problem of wanting to show with granularity patterns in the 1850 census, but lacking specific addresses. The enumerators in 1850 weren't recording and in many cases specific house addresses as we think of them today didn't exist in 1850 city, but we do have other sources to draw from. So as Leah detailed in our section of the presentation, we can work with digitized city directory records which are not a comprehensive listing, they are a biased record, but they do provide individual first and last name and address. And so from there, first matching on first and last name to first and last names in the census isn't going to get you a lot of specificity. So that's where different conceptions of like a fuzzy approach come in where we consider the enumerator pathways, proximity, the modeling the route that enumerators would have taken from census record to census record which indexes proximity in physical space. If you're near someone on a census record, you're near someone on your block, more likely based on how enumerators walk the city. So we take all those things into account in that fuzzy matching approach. Something we're going to be writing a lot more about too. Yeah, okay. So to address Mark Stern and Ben Schmidt, the question is, can you talk more about how your application would link to IPOOMS data and about the restrictive licenses of IPOOMS? I think this is something we struggled with for a very long time. And Leah, right, we'll speak to this. Yeah, absolutely. So we really appreciated all the help we've gotten from the Minnesota Population Center and IPOOMS more broadly. Because of the nature of the data sharing license, we can't have names distributed or in the map itself. But what we can do is provide a unique ID that can then be joined back to that individual account. So if you're from a research center and can go in and get a data sharing license with Minnesota Population Center, you can get the entire data set. If you're not from a research center, you can go to FamilySearch or one of these other locations and get the address of someone, one of your relatives or someone of interest. And then come to our map and look at the population, the demographics and all these characteristics at that historically accurate address and look at these patterns surrounding that area. So it's less than ideal. We'd love to be able to include names and everything in our map, but it's what we're working with and we're excited about what we've been able to do with the Minnesota Population Center and with the data sets that they've put together. So, Lynn asks about will people be able to upload their own case studies to our map? And Laura is going to address that. Yeah, we'll submit groupings of information for more public access. So as we said, the data will be downloadable for use in other formats, but we're very excited about the case studies part of the website and we really do invite participation. We're not sure yet whether you'll be able to just simply upload. It'll be slightly more curated, but once you've mastered how to use the map, you'll probably have mastered how to make a case study as well. So if you are interested in expanding the work and creating case studies, we really invite your participation. Does any of you want to add to that? I think making case studies is a great project for a class. So a class can work in small groups or a seminar as a whole and create a case study. So we would really welcome that and we would love to work with teachers about how to have them on the website. So we will figure all that out. So all the knowledge that's created by everybody out there can be shared with everybody out there. I'm gonna answer one question. It's about in 1873, a lower portion of Westchester was annexed into New York City that as we expand the project in our next three-year phase, we are gonna be including all five boroughs, including the Bronx, which is that's what it is included as, and another portion in 1895, yes. So those will be included in our next phase. It is not available right now, but in the future, when we do NASA and Suffolk, we will also be doing all five boroughs, including the Bronx. Can I just add to that? We're really excited about the next phase of this project, which the Gardner Foundation has generously refunded us for. And as has been mentioned, we're going to be doing 1820 to 1940 for the whole city and for NASA and Suffolk. And what this will add in terms of content is first going backwards, we'll be able to see hopefully the enslaved black population in New York because slavery was not abolished until later in the 1820s and newly freed African-Americans. And moving forward, we'll be able to see the great migrations of people from Puerto Rico and the American South to New York City. So we think this will add a whole new dimension for research and teaching. So hopefully we'll see you in three years with that expanded map. Yeah, do you want to? Oh, I just want to address a question. I just want to, there's two questions about expansion to other cities as well. So we are funded to do New York City. I don't need to remind people the importance of New York City in the history of the United States. The Gardner Foundation is only interested in the history of New York City and NASA and Suffolk County. But the question about comparisons is very important. And I think Wright can speak to this. Indeed, his experience is not prior to this with New York City, but was it easier to visualize New York City than other cities such as Chicago, Los Angeles regarding the structure of the city and grid and blocks versus Wright can just speak for a moment on the city he has worked on. But I'd also direct you to John Logan at Brown University. And he was very generous as well in sharing the work that he's done. He's looked at 39 different cities in 1880 and built street grids to map those out. New York was more easier than New Orleans, which I worked on previously and that it had a gridded system and a pretty standardized address system that didn't change much between 1880 and 1910. But as far as other cities go, it's really a limitation of sources. So Los Angeles doesn't really pop until 1910s and onward. So moving this far back in cities like that will certainly present other challenges. But one of the things we're really excited about with the city directory method in matching these census records that don't have addresses is being able to expand that method to other cities, not us doing the work, but sharing those methods with other researchers so that they can start doting backwards and building these other cities. And so I'd love to see all the cities in the US mapped all the towns and everything else. And that's what we're trying to set up a framework that will allow other researchers to pick up where we'll leave off and after we're long gone, hopefully we'll get to that nationwide micro level mapping. This is a question I think Dan will answer. We had a question of how we built on the New York Public Library's mapping and how many ways this project was so much built on that project. So if you could just speak a little bit about that. Yeah, and if you're going, Leah, want to jump in on this too, I know they're very familiar with that project. Without that work, we wouldn't be able to do what we've done here. It's the digitization, the crowdsource digitization of all of the maps that comprise the map worker projects at NYPL. It's the innovative methodologies and approaches towards mapping city directory records that allows us to get our 1850 data on the map. And it's, yeah, I mean, are there other aspects of it that you can think of that we're building on? The thing that I would emphasize is that all of these large scale spatial history projects depend on complementarities and collaborations with other projects. We couldn't have done this without PPUMS. We couldn't have done this without the New York Public Library's map worker and space time directory and probably a similar project like this maybe on Philadelphia or Chicago and so we'll find similar kind of challenges and we'll need similar kind of resources. So I think the opportunity is huge with libraries like the NYPL, but in Chicago or in Philadelphia or in other places to put out their resources on which similar things can work for another city. And we really could kind of remap, re-understand urban history at this micro level which is completely unprecedented. The only thing I would add to that along the same vein of the cumulative nature of the work is that in the same way that this project could not have been done without each of our disciplinary perspectives and expertise. It's not only the building on the resources that others have produced in terms of source material but also building on the methodologies that others have pioneered. So whether somebody wants to take some of our approaches and start to build on those for New York or for completely other places. Those are approaches, techniques, frameworks and methods that are right for the taking. And we decided in the very beginning that we would work from a principle of transparency so that all of our methodologies will be available to anybody who wants to use them and learn from them. Just a quick addition because we might not have actually showed it on the slides that we showed today but the historical maps are in the background so you can switch on the historical layers in relation to the dot density maps and the digital street grids that you saw. So the historical, we've collated actually all of those maps in the background of our map. And as soon as we publish the link which we should put in the chat at the right now you will be able to see that you can utilize those. One of the other things I'm really excited about is when we were building these historical street grids we had students when they found an issue or sources that didn't line up we had them document those pretty rigorously along with the latitude and longitude. And so one of the things that we're excited to produce in the future is a map that shows our certainty or uncertainty across our street network and across the time periods in the city. That answers, I just want to give recognition. We have questions about workflow, how to employ students to do this, about training of students and graduate students. I'm not sure we're gonna get to all of those and we will answer you individually because I do see these questions but I wanna just speak about, there are several questions about understanding commercial real estate, industrial zones, how they appear in our map and I'm gonna open it up to others to answer but I want us to remember that we are based on census material. So in earlier in the early 19th century, people often lived in their place of production so that if the census taker confined a person living there which in many ways we capture change over time in the way that the residential space and industrial space or even pre-induct spaces of work have happened but I'd like other people to relate to that also. I'll just point out that as you start playing around we know we have a lot of layers available but among those in the historical map face layers are geo-referenced fire insurance atlases for each of the time periods and those give an indication of land use of course. So those are available for comparison many of which like the Paris Fire Insurance Atlas of the 1852 to 1854 was digitized by the New York Public Library and so that data as GIS data or as images are also available. And I would add one more thing that we don't have land use maps specifically built by us but just as Redeca mentioned when you are actually mapping occupation that is oftentimes a clue for land use. Laborers along the shore or port workers or garment workers, workers in the needle trades that's usually homework, sweatshop work and so on. So while it's not direct land use coding obviously nevertheless occupation oftentimes can serve as some sort of an understanding of what kind of activities what kind of industrial or commercial or other activities take place where exactly in the city. And of course this is one of those points that we can emphasize that. And next step for another project or for someone else interested in this could be to work with the city directories which do actually have more of that kind of land use information, work with our data set, combine them and try to explore land use in relationship to census data, residential information and so on and so forth. So I have, we're gonna end with a meta question but a very specific question has been asked is how are we protecting this site for posterity as we all know NYPL has taken down map warper. So Leah would you like to, would someone like to say how we've protected it? Yeah, Laura. That goes to Laura. Yes. We actually got a subsidiary grant based on this project where we worked with many people in Columbia libraries and a much broader group to understand how that's what I said at the beginning of the presentation how to curate and preserve the data. So the data itself will be preserved in the Columbia library and the map interface will of course be vulnerable to the vagaries of how the internet evolves and to various softwares, et cetera. But the data at the behind our map is safe for posterity. I believe that. So this is the final question and please, as I said, this is a conversation. We look forward, you can find all of us on the Columbia website if you have specific questions about methodology, about data, about larger questions, about where we're going, please feel free to email us. But I'm gonna quote Ben directly and I'd like each of us to answer this. I'd like to spin a meta question back to the team. Obviously this project has much to teach us about historic New York and we can all look forward to digging in and testing hypotheses. But I'm interested in what this project has taught you about the process of mapping more broadly. Are there ways this work has challenged your assumptions or led you to best new practices? Anyone wanna start? It's a big question. Well, I'll start. So I'm gonna now just say the reason I took the Mellon Bootcamp and how many, everything I knew about mapping was taught to me by Lou. So yes. So I took it cause I teach, of course, Immigrant New York and there are many, I would call them MIPS about neighborhoods in New York and their uniformity for different groups, from immigrant groups from different places which we always know is more complex. And I wanted to understand how I can show my students the complexity of what it actually was like to be an immigrant in New York when you were not surrounded by everyone who spoke the language of your birthplace or were you? So I would say that for me, it has shown the power of thinking about how New York has changed over time but I have become fascinated by Brooklyn. Brooklyn is fascinating to think about cause it is often an area of second settlement. I've never thought about it cause the way we write New York City history sometimes does not pay attention to Brooklyn. Brooklyn and Manhattan are part of one unit and it has really helped me think about how groups move through the city across time changing many different neighborhoods and thinking through that and how we write the history of how immigrants interact and change cities and cities interact with them. Yeah. It's maybe it's a cheat of an answer. I don't know that the project, I mean, it challenged us in, I'll speak for myself, it challenged me creatively, especially in many different ways but I don't know that it challenged my assumptions. If anything, I think it confirmed my assumption or the operating premise that with GIS work good enough is never good enough and the whole process, all of our iterative sort of refinement process was an absolute joy to continue to solve harder, exponentially harder and harder mapping problems because we knew we could get it better. The other sort of confirmed assumption is that the city is a spatiotemporal entity and the sort of the conceptual frameworks and the methodologies of spatial analysis need historical analysis and vice versa. Well, at one level, I think that what our team did with all the students and our collaborators was something magical, right? That they were able to assemble and put together all these data to create these maps. But for me, it also confirmed an understanding that all of this is a representation. It's not necessarily what was really there and it teaches us how space is organized. It doesn't just automatically create itself, it's organized by people with interests and so how that space got created and how that space changed and who got counted in the census. All those are in some ways arbitrary measures that humans created to explain their environment. So I think while we captured what they were doing, I think it also teaches us that mapping and census taking are in some ways very arbitrary exercises. I would like to say that the most important lesson I learned from this is the extraordinary power of microdata, just to be clear, individual level of data. I do believe strongly that the future of social science research and social science history is based in microdata, individual level of data that allows us to kind of examine social phenomena at the individual level but also at large scale and the point about that though is that to deal with 6.5 million census records plus then adding 1914 New York City data for like five million people and the number keeps growing. That means that these kind of projects which I believe are the future of social science history type of projects are team projects, collaborative projects and that is the biggest payoff to working at team, to learn from each other and to really try to pool each other's skills and assets as components. So those are my two big takeaways. Yeah, the microdata is just a fascinating resource. And one of my big takeaways from the project is the cultural importance of making these projects public. There's so many big mapping projects and big data projects that come out of academia that never see the light of day in the public. And so by unlocking these data sets and putting them in the public's hands and everyone else's hands, it's just invaluable resource that I think we don't do enough of. And a big part of it is just the amount of work that goes into that getting it online, getting an interface. And so I want to applaud Dan and Stamen and everyone else in really coming up with an intuitive math interface and getting it online and getting it user friendly. And just, it's so exciting to be at this point and get to see it out there in the world. Look forward, all of you and your students will do with it. Laura, do you want Dan here? What May said really resonates with me, I feel very immersed in right now just in like the tunnel vision of census practices. Gearing out this project while the 2020 census was unfolding was troubling and illuminating and just seeing the resonating with the 1880 census with the 1910 census that we've been working so closely with and quite literally following census takers around as they conducted their work has made me just think about a lot of more about data collection practices and how it is in processes. That's what I'm saying. But I think in terms of the new best practices, I think we're in such a privileged space over here to have had this kind of funding to do what we're doing. So I think what we're seeing is the extraordinary amount of work that it takes to put something like this together. I never thought I would be involved in a project as detailed, as comprehensive as this one is and especially in terms of its collaboration. I think it's really interesting what you're saying, right? But you're happy about the public-facing nature of it because I think so much of academic history ends up in peer-reviewed papers and not public-facing things like this. But I don't think we could call this a best practice. I think that it's incredibly rigorous work that we've learned how to do it through the process of doing it only because we have these amazing grants to support this work. So everything is public-facing. All our code will be public. Anybody will be able to replicate the work that we've done in another city, but we understand that it's not so much about best practices. It's just about extraordinary detailed work that goes into making something like this. So I thank everyone on that. I think we're gonna close this launch, but I wanna close by thanking the Gardner Foundation because what Laura said, the detailed work and the number of students involved to make this possible for what we've shared with you today and we look forward to what you do with it would not have been possible without their generosity. They have been committed to supporting the study of the history of the city of New York broadly conceived for a long time and they have made this possible and we thank them. And we thank everyone. I think everyone on the team has learned so much about thinking about how we think about the history, the past and the present. As Dan said, we went through the 2020 census and all the politics surrounding it, thinking about past censuses who was left in, who was left out. We've really, really learned a lot from each other and we look forward to learning from all of you who are gonna use this tool and teach us and show us new ways that we can think about the development and history of the city of New York. So thank you so much. No, so thank you to the attendees for staying. I think the fact that so many of you are still online is testament to the fact that you're really interested in how we're doing what we're doing and will help us in moving it forward with the census. So we really do look forward to collaboration with other institutions, with other academics, with other universities, with other high schools. That's why we made this map. And last but not least, we'd like to thank the GSAP behind the scenes, IT people who made it impossible when we decided we wanted to do our launch all in the same room, like the way we've been working on this project for three years, which is a challenge in this moment. That's why you have to enjoy us with our masks on. But we'd like to thank them for making this all possible. And thank you everyone for coming and we look forward to your questions and everything you have to teach us about this tool that we've now put forth in the public. Bye.