In Introduction to Art Image Access, Patricia Harpring states, “the reason for designing appropriate data structures and devoting considerable time and labor to indexing
subjects in visual works is to provide good search and retrieval for the images being catalogued or indexed” (13). With this in mind, I decided to evaluate a library collection and a museum collection, as a user might, by searching for specific images.
I wanted to find the photograph below of a man suspended high above the New York skyline. Unlike some patrons, I had a lot of information to seed my search: I knew the photograph was taken by Lewis Wickes Hine during the construction of the Empire State Building and is part of the New York Public Library’s (NYPL) collection.
Title: lcarus, high up on Empire State
Photographer: Lewis Wickes Hine
URL: http://digitalcollections.nypl.org/items/510d47d9-a92b-a3d9-e040-e00a18064a99
The NYPL Digital Gallery offers a keyword search, but a search for “Lewis Wickes Hine” returned over 700 photographs. After clicking around the site, I was able to find the collection guides, which are finding aids for the various collections. The Lewis Wickes Hine photographs belong to the Lewis Wickes Hine: Documentary Photographs, 1905-1938 collection, so determining the correct collection was rather easy. The collection itself is broken down into five different sections, with no description other than their titles. Because I knew where the photograph was taken, I knew that the section on Photographs of the Empire State Building under construction was the right place. The section contains 47 photographs, which I had to browse through to find Icarus.
The catalog record for the photo above contains the name of the photographer and the title, along with the note that a cataloger devised the title. The subject headings are:
- Empire State Building (New York, N.Y.)
- Building construction -- New York (State) -- New York
- Skyscrapers -- New York (State) -- New York
- Steel industry
Upon reading the subject headings, I scrolled back up to verify that I was looking at the correct record. But yes, the cataloger had somehow missed the fact there is a person in the photograph! While all of the subject headings are correct, the LCSH term Iron and steel workers would reflect the main topic of the photograph.
Upon further inspection, I discovered that all the records for the photographs in this sub-collection are almost identical. For example, the following two images have nearly identical records; only the titles and locator numbers differ.
Title: View of the building rising to about sixteen stories
Photographer: Lewis Wickes Hine
URL: http://digitalcollections.nypl.org/items/510d47d9-a941-a3d9-e040-e00a18064a99
Title: A worker riding on a crane hook
Photographer: Lewis Wickes Hine
The similarity of the subject terms suggests that the collection was not cataloged at the item level. My successful search would seem to support Jackie M. Dooley’s assertion in Processing and Cataloging Archival Photograph Collections that we do not need to catalog at the item-level instead using group-level descriptions to guide users (89). However, I disagree. I got to the Lewis Wickes Hine collection based on a great deal of knowledge of the image I was seeking. Imagine for a moment, that all I knew was I was looking for a Depression-era photograph of a steelworker high above the skyline. I might have started with searches like “steelworker” or “construction worker”, browsed through hundreds of photographs, and still not found the correct photograph!
While the photographs in the collection are related, they are not “about” the same thing and should not have identical subject terms. The visceral reaction that these descriptions are missing something is backed by the method the Categories for the Description of Works of Art (CDWA) uses for analyzing images, which uses Erwin Panofsky’s three level approach in which the cataloger describes the generic elements, identifies the iconography, and interprets the major themes in the image (Harpring, 3).
Had the cataloger taken this three-level approach, the records could not be identical, the preceding image (and the image of Icarus) would contain references to steelworkers or at least, people, to satisfy level one cataloging. The individual catalog records would allow users with limited knowledge to still have something useful to begin their searches.
My investigation of a museum’s cataloging methods was more frustrating. I began with a search for the image of John F. Kennedy, Jr. playing under his father’s desk in the oval office.
Title: JFK with John, Jr.under Resolute Desk
Photographer: Stanley Tretick
URL: http://store.jfklibrary.org/PAAAAALOHMDFNOEM/Product
The John F. Kennedy Presidential Library and Museum allows users to use traditional finding aids to peruse the collections or users may search by a variety of metadata fields including photographer, persons depicted, places, date, and many more.
I began my search using the traditional finding aid to test its effectiveness. Each finding aid describes the contents of the collection at a high level and lists the individual folders or images by title. Folders have very little information other than a somewhat limited title like White House rooms and the date the photographs were taken. The White House Photographs collection contains 46,500 images. It seemed unlikely I would find the desired image simply by browsing.
The finding aid does allow users to refine their search based on a variety of metadata fields, so I was able to limit my search to photographs containing John F. Kennedy, Jr, limiting my results to 137 images to browse. One of these images is of father and son on vacation shortly before the president’s assassination.
Title: President John F. Kennedy and John F. Kennedy, Jr., in Newport Rhode Island
Photographer: Robert L. Knudsen
URL: http://www.jfklibrary.org/Asset-Viewer/Archives/JFKWHP-KN-C30005.aspx
The individual record contains a great deal of information, including separate fields for the photographer’s authoritative name, the authoritative names for each person depicted, the date the photograph was taken, and the authoritative name for Newport, Rhode Island as well as a broader, separate entry for Rhode Island. The subject terms include “Boats and boating”, “Kennedy family”, and “Voyages and travel”.
The record reflects Panofsky’s three levels of description. For level one, we know what is generally in the photograph: John F. Kennedy, John F. Kennedy, Jr. and a boat. For level two, we include broader ideas: the photograph is not of Newport, Rhode Island or the Kennedy family as a whole, but these are important access points for describing the photo. Finally, for level three, we get “voyages and travel” so that users looking for the idea of the Kennedy’s on vacation have an access point. It is a well-crafted record, giving users many access points to trigger their searches.
Yet, after browsing all 137 photographs, I still could not find the photograph of JFK, Jr. under the desk. I searched across all collections, but still no luck. I started to wonder if the photograph was part of the museum’s collection, so I Googled “jfk jr. under oval office desk”. The second result was to the photograph for sale in the museum’s online store. I suspect that the image is restricted by copyright and cannot be shown on the museum’s website, but it is still thoroughly depressing that all of the cataloging effort shown on the museum’s site did not get me what I wanted and I had to browse through hundreds of photographs fruitlessly.
Dooley, Jackie M. Processing and Cataloging of Archival Photograph Collections. Visual Resources XI .
Harpring, Patricia. The Language of Images: Enhancing Access to Images by Applying Metadata. <http://www.getty.edu/research/publications/electronic_publications/intro_aia/harpring.html>
No comments:
Post a Comment