My term-long project is the technical foundation of my primary research for my dissertation, which will be examining political cartoons as either representative/reactionary or causal of society’s interaction with nucelar fear and anxiety.
I need to build a database (most likely housed in Omeka…sorry if that terminology is incorrect) where I can do complex queries on these single panel cartoons. I’m starting by examining the corpus of Washington Post cartoonist, Herblock, for the years 1960-1974, but for my dissertation will be expanding to examine at least two other cartoonists, Frank Miller of the Des Moines Register and Paul Conrad from the Denver Post and the Los Angeles Times.
Herblock’s cartoons in particular are very text heavy. There is generally a one line caption that is the “joke” of the cartoon, but the drawing itself also tends to be heavily commented. Herblock extensively used labels because he tended to draw the same “stock” character and just label them as the situation deemed necessary. So the large-bellied man could either be Congress as a whole, a specific Congressman or any other agency/entity/individual that seemed to fit that description.
Beyond this, there are many other characters that make a routine appearance in his cartoons: Uncle Sam, an anthropomorphized atomic bomb and H-bomb, John Q. Public, and any president or other world leader. In addition, many cartoons deal with specific themes, topics, or events that I like to include in the metadata like; McCarthyism, the Vietnam War, Civil Rights, Education, and various presidential elections. This is a lot of metadata that right now I can get to visually by looking and reading each individual cartoon.
Because I am eventually going to be looking at thousands of cartoons for my research, I’d like to be able to get at this metadata much more easily. I would like to be able to search for all the cartoons featuring both Nixon AND McCarthy in a particular date range but are NOT on the theme of McCarthyism, for instance, and have the relevant cartoons appear. My aim is to figure out the best workflow for getting these cartoons, (right now for Herblock in a large PDF document with a two page spread of the cartoons) OCRed, and/or tagged, and get that information into an Omeka database as easily and painlessly as possible.
I know that this project is going to be heavy on the manual labor and no amount of skills/technology will allow me to fully automate this process. I would just like to avoid creating a new image file for each cartoon and then manually coding the metadata I want attached to each of them. I’m leaning towards an Omeka based project because it already allows for some of those advanced search options I want, as long as I have the relevant metadata fields created and populated.
For my final project I know I will not be able to even get all of Herblock’s cartoons for my proposed date range into the database. I hope to get enough that I can start examining them to see if there are any prominent themes/issues that he was concerned with over time as an individual cartoonist, but my dissertation is focusing on representations of nuclear fear and anxiety so I hope to be able to begin parsing out some of the prevalent topics he was concerned about. As this happens, I hope to create an Omeka exhibit highlighting these topics and placing them within the larger historical framework of nuclear culture and politics.
Here’s an example of one of Herblock’s cartoons to illustrate what I mean.
Caption: The Sorcerer’s Apprentice
Date: October 7, 1990
People: Ronald Reagan, George W. Bush
Keywords: Voodoo Economics, Budget Cookery
Themes: Economy, Budget
Description: George W. Bush dressed as a wizard places books entitled “Budget Cookery” into a cauldren labelled “Voodoo Economics” and looks frightened as it explodes, while a painting of sorcerer Ronald Reagan hangs in the back ground.
I’m still debating whether I want to put in the effort of also including this 1-2 sentence description of the cartoon. This may be worth the effort in the long run not only for my own research but it will allow my database to be searched by other scholars as long as I keep the images of the cartoons hidden as there are copyright issues. But this would be an entirely manual effort.