This week, my LIS 7850 Digital Libraries class participated in a project designed to “introduce us to the issues surrounding access to photographic images through basic descriptive entries.”
The exercise involved creating metadata for images from the Virtual Motor City Project, a large collection of historic Detroit area photographs donated (both in physicality and in copyrights) by the Detroit News to Wayne State University Reuther Library in 1997.
Each class member was assigned several images, by ID number, and asked to fill in missing metadata in the following fields: title, description, free text tags, and Library of Congress subject headings. The images had a few existing bits of metadata, which I understood to be notes from the outside of the each original envelope in which the photos were kept by the Detroit News.
(I could easily picture this, having participated in the acquisition of a newspaper photo morgue from the Xenia Gazette a few years ago, when it was acquired by the Greene County Room, the local history department at the Greene County (OH) Public Library. The acquisition consisted of thousands of little folders and envelopes with a label and a few notes on the outside of each one.)
I was assigned 30 images to describe. Unfortunately, 19 of those ID numbers did not currently have a digital image attached to them, so I could not see the picture. I was told to skip these, which makes sense because it’s pretty hard to accurately describe something you can’t see, when all you have to go on are a few (sometimes conflicting) keyword notes. The other 11 images mostly depicted Catholic Churches in Detroit.
So I have the image to look at and the name of the church from the original metadata. Now all I have to do is write meaningful descriptions for them. Sounds simple enough, doesn’t it? Well, perhaps not for everyone. But for me, it should have been. I have worked in public library local history departments for over 4 years, so I do this type of research pretty much all the time.
And yet there was one serious challenge that made me feel like a fish out of water. This is an online course at Wayne State University, which is in Detroit, so it makes sense that any local history project we do would be about Detroit. However, I live 200 miles away. And most local history resources tend to be one-of-a-kind (or nearly so) and only available in person at a library or repository that is geographically near the place which they are about. You see where I’m going with this. Two hundred miles (about 4 hours) is a bit of a far drive to make a run to the library to check the old city directories, city maps, or city/county histories that my fingers were just itching to reach for. If this project had been about old buildings in Dayton, Ohio, I would have been golden, as those resources are at my fingertips every day.
So I had to improvise, using sources I was able to find on the Internet. I regret that in some cases, these sources may not have been as complete or as officially “reliable” as I would have preferred, but it’s all I had to go on. The sources I used varied from place to place, depending on what I was searching for exactly. But I did use a few resources more than once:
o Google Maps street view: http://maps.google.com.
o Archdiocese of Detroit web site: http://www.aodonline.org/. (Most churches had individual web sites with history and other information.)
o Flickr – http://www.flickr.com. (I believe that most people who bother to go out looking for historic buildings to photograph tend to label these things correctly. They often provided other information that I considered “clues” to be followed up on.)
o Google Books – http://books.google.com. (I found a Burton’s 1922 history of Detroit there.)
o Atdetroit web site / forums – http://atdetroit.net/. (Discussions about historic buildings often included additional photos, information, and sources. Again, these provided additional “clues” to be verified.)
Some of these less “reliable” sources, such as Flickr or the forums, were particularly helpful in identifying churches whose names had changed or that no longer existed.
For instance, the church identified as Beulah Baptist Church (ID #20252) is now known as Allen Temple. I might never have found that, if not for pinehurst19475’s Flickr photo, which stated this information and clearly depicted the same building.
Or, in another example, I was just about to give up on Our Lady of Sorrows (ID #20258). (Google was giving me nothing.) But then I noticed a sidebar list of “portfolios” containing this image, and where most of the images were only included in one portfolio (the one I was working out of), this one had two: the other was called “Our Lady of Sorrows Church Fire.” This portfolio had a few other images, showing the church ablaze, and some of the other metadata notes mentioned a Briggs plant fire. I searched the Internet for that. I came across an Atdetroit forum discussion about the fire that consumed the Briggs plant and the church, where members had posted several more pictures (many of which had come from the Archdiocese archives) and included a date for the fire [April 10, 1963]. This should of course be verified—I would probably start with old newspapers on microfilm since the date is exact; seems like the simplest search route, but again, unfortunately I am 200 miles away!
There were a few other interesting searches similar to these, but I think I will spare you the details of my research on each individual image.
I think I managed to hold my own in writing the descriptions, even if I did feel quite a handicap in the resources I had available. I hope that these descriptions will meet the expectations of the project manager and the future users.
The other main aspect of the metadata we were asked to provide was to add Library of Congress subject headings and free text tags to the images, in addition to titles and descriptions. Tags are easy. Tags can be anything you want (as long as it makes sense). I was a little worried about figuring out the LC headings. I’ve taken the cataloging and classification class, but LCSH is still intimidating to me. The professor gave us several pages of example LC headings, and that helped a lot. Having so many similar images made it easy, too. Once I had figured out the appropriate headings for the first church, it was easy to “plug in” very similar headings for all the others.
This gives me more confidence for adding LC headings to the Dayton Metro Library digital collections that I manage. Up until this point, any LCSH metadata that was added was created by one of the cataloging librarians. Unfortunately, I recently found out that the cataloging librarian we had been working with (who was actually a Wayne State grad, I found out recently – small world) accepted a new position elsewhere. I had wondered who would help us next. Perhaps I will take a stab at the LCSH myself. Maybe it’s not so bad after all…