Technical Failure
“We stumbled onto digital technology for oral history under the false assumption that it would save us money and personnel in the long run.”
– Williem Scheider (p. 19).
Even my solutions look exactly the same.
The story of access to oral history is the story of technology and data. Oral histories like any other archival or historical material is a form of data, only the difference between a painting and an oral history recording is that oral histories having become increasingly dependant on advancements in digital technology, in particular its ability to be access. The history of oral history access works as a case study for technology as it reveals opportunities, assumptions, and mistakes which humanity repeatedly makes when it comes to technologies. Most significantly it reveals how the question “after the revolution, who’s going to pick up the garbage on Monday morning?” is ignored. This project aims to break this vicious cycle by focussing on maintenance, but first I must elaborate on why this focus on technology has led to the vicious cycle in the first place.
The access to oral history recordings has historically been framed as an issue with its medium and how it obstructs its usability. The audio format is continuously compared to the infinitely more searchable medium of text. During the late twentieth and early twenty-first centuries oral historians started to experiment with various new technologies in an effort to make the searching of the audio easier. The resulting solutions generally came in two formats: the first were personalised curated collections of oral history recordings and the second were softwares to help navigate oral history recordings. The former includes such projects as Project Jukebox (year), VOAHA (year), and Civil Rights Movement in Kentucky Oral History Project Digital Media Database (year) which were described by Doug Boyd as well-curated exhibitions rather than archives (Boyd, 2014). The latter included Interclipper, a software most prominently championed by Michael Frisch, and Stories Matters from Concordia university. Sadly, none of these exist anymore. My research into these endeavours revealed how an assumption that technology will fix things results in the maintenance of these technologies being ignored, resulting in many failed and obsolete projects. I wish to dive further into this period to justify why the subject of this project specifically is the maintenance of access to oral histories and I avoided the allure of the technical fix.
The central focus of the majority of these endeavours was reinserting orality into the oral history archival process [ref: Boyd p. 86]. This was achieved by syncing transcripts, key words, and additional meta-data to the relevant points in the audio files. This is now a common technology but at the time it was cutting edge. However, what these project missed was how these technologies would be maintained and how they could be adopted by others.
Many of the projects were not maintained. Kentucky was digitally abandoned (Boyd). VOAHA slowly fell apart and had to absorbed into the main library and cannot be added to (Gluck). Stories Matters cannot be downloaded (image). I have found no trace of interclipper. Project Jukebox does still exist having moved on from the CD days. Overall it seems many underestimated the worked needed to keep things going.
The failure to be adopted by other oral historians is not surprising in the cases of projects which were more digitally curated collections such as VOAHA, Project Jukebox, and the Kentucky oral history project. They were tailored for their particular configuration of the oral history problem, not for wider adoption. [give examples] The softwares should have been more adoptable by other projects yet this turn out not to be the case. Sherna Berger Gluck opted not to use Interclipper in the development of VOAHA instead developing her own system because Interclipper did not supply a digital file of the entire oral history recording alongside the clips, and its database was not compatible with the internet (Gluck, 2014). During the research period of Stories Matters students at Concordia University also tested Interclipper and found it frustrating to use, worried about certain information being lost, and thought it too expensive (Jessee, Zambrzycki, and High, 2011; High, 2010). It is unclear whether Stories Matters itself was used by any other project although it is no longer downloadable from the developer, Kamicode, website.
One oral history technology that has been successful is Oral History Meta-data Synchroniser or OHMS, a web-based software developed by Doug Boyd after his work on the Civil Rights Movement in Kentucky Oral History Project Digital Media Database (reference). OHMS links up a transcript and keywords to a recording, allowing the user to navigate the recording by searching the text. This text, including the keywords, are created by human labour most often students who type everything out (reference). It definitely benefitted from being developed a bit later on when internet technology was more establishedand has now been adopted more widely. [something about its relationship with Omeka and its addition to a software suite]. Overall OHMS has been the closest to solving the particular formulation of the oral history problem which focussed on the assumed inaccessibility of the audio recordings. However it has struggled to be adopted in Europe due to GDPR restrictions and has not necessarily reduced the amount of labour needed to archive oral histories.
Back to Frisch
“Oral history libraries are closer than most archivist want to admit to that shoe-box of unviewed home-video cassettes in so many families—precious documentation that is inaccessible and generally unlistened to and unwatched.” [Frisch, p. 223]
The quote used at the top of this section came from William Scheider who worked on the development of the original Project Jukebox. Later in the book Oral History and Digital Humanities published in 2014 Scheider reflected on the project and concluded the reliance on technology might have been a mistake. To a certain extend I believe his analysis to be correct, the projects I have mentioned would fit into the category of “technical fixes” [proper definition of technical fixes and how the focus on technical solutions can create ‘unforeseen consequences’]. Technical fixes would fall within Ukeles’ development system. They are “pure individual creation[s]” focussed on “the new; change; progress”. However, this does not mean I should ignore technology, because its use is unavoidable and standard in accessing oral history archives is standard and cannot be avoided, instead my predecessors, like Morgan in his blog post, indicate a necessary change in perspective which focusses on the maintenance of oral history recordings.
We are standing on the precipice of a new technological age with the continuing rise of Artificial Intelligence (AI). This has not gone unnoticed by the oral history community as the American Oral History Association’s [annual conference in 2024 had the theme of Oral history and AI.] AI, especially the Large Language Modules and the chatbots, have opened up a whole new world of access to all information and data, including oral histories.
ChatGPT, which is considered the catalyse of this AI arms race, was release halfway through my project in the winter of 2022, but it did not result in me pivoting towards a more AI centred project. The reason I did not jump on this undoubtedly world changing technology is because of what happened with oral history during a similar technological revolution in the late twentieth and early twenty-first century.