Millions of books have been digitized since 2000. When we want to discover them, though, researchers often have to choose between two often unsatisfactory choices: keyword search that dips into the library at random, and historical classifications that carry the century-old priorities. I'll explore how new techniques in visualization and machine learning make it possible to better understand the elisions, biases, and priorities that shaped taxonomies of knowledge as the modern research library emerged. What groups, people, and classes were consciously or unconsciously excluded from the organization of knowledge as the state took over America's library infrastructure after 1900? Tools from the digital humanities can help us draw a new picture of the politics of knowledge at the turn of the 20th century, and let us better understand the digital source base that is essential to humanities scholarship in the 21st century.

 

Come early to join us for a pre-talk reception, which will begin at 2:30pm.