Walk the streets of any U.S. city today, and you might come across historic markers or masonry etchings indicating what the buildings used to be. It is always fascinating to learn what our neighborhoods, cities, and towns used to be —factories turned residences, street names changed, the places and spaces our predecessors lived, ate, and worked? The Sanborn Maps Navigator, a new tool created by a Library of Congress Junior Fellow Selena Qian, allows you to travel to familiar and far-off neighborhoods to explore the past.
Qian completed the project from afar in 2020, as a Junior Fellow at the Library. Each summer, the Library welcomes a cohort to its Junior Fellows Summer Internship Program. This year, due to the COVID-19 crisis, Junior Fellows worked entirely remotely—a challenge, but one that allowed dozens of new explorations of content, services, and access like Qian’s.
As Qian explained in July, her project grew out of a Junior Fellow project in the Digital Strategy Division aimed at showing experimental ways to share digital content at the Library. It mashes up two collections created for entirely different purposes: 32,000 atlases from within the Sanborn Fire Insurance Atlases collection and 1,494,585 photographs in the photographs culled by the Newspaper Navigator machine learning experiment. The result, called the Sanborn Maps Navigator, provides a prototype for how digitized content and new technology can lead us in unexpected directions.
Beginning in the 1860s, the Sanborn Company created atlases full of maps designed for use by fire insurance companies and underwriters. The maps depict cities in the United States—with some from Canada, Cuba, and Mexico—and their streets, buildings, and man-made infrastructure. The maps contain an astounding level of detail about building materials, measurements, construction, and other elements of the neighborhoods—needed by insurance companies, but rarely recorded elsewhere as comprehensively. This work went on for decades, and the Library now holds over 50,000 of these historical atlases, around 35,000 of which we have made available online (totaling 438,313 map sheets!).
But that’s not all the Sanborn Maps Navigator offers. Qian recognized another incredibly valuable source of visual information about places across the United States in the past: the Newspaper Navigator dataset. Library Innovator in Residence Ben Lee used experimental machine learning technology to separate visual content from over 16 million newspaper pages in the Chronicling America historic American newspapers site. The newsy portraits and photographs of weddings, hotels, food, and more add to the sense of place depicted in the maps.
Qian’s experimental prototype project ties together place and life in an interactive map visualization tool. Created in only a few weeks, Qian’s Sanborn Maps Navigator allows you to click the map to find a particular location, or choose randomly by clicking “Surprise Me!” Maps and newspaper photographs will update as you change geographic locations. “I wanted to create something that would draw people in,” Qian explained, “and encourage them to explore the information in an intuitive way. I also wanted to add some element of discovery and serendipity, by tying in the historic context of the newspaper headlines and images.”
Learn more about how Qian accessed the data and built the tool in the site’s Frequently Asked Questions page, and more about using the Sanborn Maps and other bulk downloads via the Library of Congress’s APIs in a recent “LC Maps for Robots” post.
Whatever you do, we hope you’ll find some time to get lost in the past with the Sanborn Maps Navigator—or find just what you’ve been looking for!
Note: this blog post has been updated with a corrected link.