NotesFAQContact Us
Search Tips
ERIC Number: EJ771619
Record Type: Journal
Publication Date: 2007-Jun-15
Pages: 1
Abstractor: ERIC
Reference Count: N/A
ISSN: ISSN-0009-5982
An Index of Horror: To Catalog an Extensive Video Archive of Holocaust Testimony, the U. of Southern California Turned to the Human Touch and a System that Could be Used for Other Large Archives
Carlson, Scott
Chronicle of Higher Education, v53 n41 pA32 Jun 2007
On video, a woman describes how her life was shattered. She speaks about her family splitting up, about her loved ones being killed, about one of the most systematic genocides in history. Indexers at the University of Southern California's Shoah Foundation Institute for Visual History and Education have watched 120,000 hours of these testimonies since 1998, paying attention to every detail and indexing each and every minute. The Southern Cal collection is one of the most comprehensively indexed video collections of its size. The indexing allows researchers to search for various topics through keywords and gain access to the testimonies at the very moments that those topics begin to be discussed. Video is a powerful medium, but as a research tool it is not much use unless it is indexed. Some people at Southern Cal see the institute's labor-intensive, manual method as a model for other video archives. The method of indexing the tapes in one-minute segments and using keywords has been patented, and the university plans to market the patents to other archives. The technology has already been licensed to the state archives of Italy, which is processing video in its holdings. Organizers of the testimonies found early on that speech-recognition technology would not work for indexing their collection. Speech-recognition software can recognize most words if a speaker is well recorded and enunciates clearly, but the technology breaks down under other circumstances. In the case of the Shoah testimonies the speakers sometimes had thick German, Polish, or Yiddish accents, and would occasionally lapse into their native languages. Some survivors did not speak English at all. As they recounted their lives before, during, and after the Holocaust, many were overcome with emotion, and speech-recognition software cannot interpret sounds like those. The software has little capacity for processing meaning in context. A Holocaust survivor may talk about having been forced to break the Jewish dietary laws, but might never say the word "kosher." He might instead say, "The Germans made us eat pork." Or he might say the same thing a hundred other ways. Although video is a rapidly growing medium of communication, the extent of the demand for manual indexing is uncertain and due to its high cost and labor intensiveness, may not be practical for many archives. Most video archives instead try to index the content of video through a combination of automated methods. They might use speech recognition together with transcripts, or closed captioning, or image processing. Multiple methods can collectively zero in on the content and recognize what's going on in the video. The Shoah Foundation Institute's manual method, however, might be among the best ways to index historically important oral histories and testimonies: as one former indexer responds when asked what he learned about humanity from the experience: "That we are designed to communicate with each other, on a physical level and an emotional level, and we are designed to unravel the meaning of otherness."
Chronicle of Higher Education. 1255 23rd Street NW Suite 700, Washington, DC 20037. Tel: 800-728-2803; e-mail:; Web site:
Publication Type: Journal Articles; Reports - Descriptive
Education Level: Higher Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Location: California