overview of case study/background of collection
review of case-study
Julian P. Kanter Political Communication Center/Archives
In the Spring of 2019, the University of Oklahoma experienced financial issues contributing to closing or cutting funding for campus collecting repositories. The Julian P. Kanter Archive once operated under the Communications Department with an impressive technological studio and lab environment named ‘The Political Communication Center’ (PCC)’1. In 1985, the University purchased the Archive from a private collector, Julian P. Kanter which contained 25,000 political commercials.
The Collection contains television and radio commercials, debates, interviews, and other political programs from 1912 to 2020 (Wood 1990)2. The Collection covers all levels of political races: presidential, U.S. senatorial, gubernatorial, and other statewide offices, congressional, state legislative, county, and municipal, judicial, school board, etc. The archive has materials from all 50 states as well as 25 international countries. The Collection also contains a number of commercials for ballot issues, propositions, advocacy commercials, and public social policy questions.
The PCC ran successfully as a unit with substantial funding for a director, full-time archival staff, student technicians, and 20-27 members of the Advisory Council for many years (1985- contributing to the Collection expanding to 56,000 commercials by 1996 (Haynes, K. J., Kaid, L. L., & Rand, C. E. 1996)3. By 2003, the Collection had grown to 70,000 cataloged items with 10,000 commercials “that have not yet been cataloged”, according to the director of PCC (University of Oklahoma November 24, 2003).
In 2012, the Collection was estimated to be 90,000-125,000 ads and reported that PCC had captured more ads by July (3,800 ads) than it did throughout the entire 2008 election (Harkins 2012; Besser and McGuire 2008). In addition, the 2012 report documented 40 boxes of commercials had “not been opened” which could account for tens of thousands of spots4.
The 2012 report marked a significant transition from collecting analog or physical media to capturing digital advertisements from the Internet. This shift was prompted by the substantial increase in the generation and publication of ads facilitated by the Internet, resulting in a rapid expansion of collection materials. Moreover, this shift brought about a transformation in the approach to gathering information.
Unlike analog materials that typically lacked extensive accompanying information and necessitated additional research, Internet metadata provided instant accessibility to relevant data. However, as the rate of content ingestion and the number of available ads continued to rise, backlogs began to accumulate in tandem.
The 2012 report was the first reference of Collection mismanagement with a statement from the lead archivist describing the collection as “every manner of media from 35mm film to VHS tapes to cassettes and DVDs all strewn about in no particular order” (Harkins 2012).
At the time of the Collection’s procurement by the Center, PCC management reported a total of approximately 119,000 ads encompassing the entire collection5. Discrepancies in the collection counts can be attributed to the ongoing growth of materials and the diminishing effectiveness of managing the existing collected materials. Furthermore, as items were digitized to fulfill patron requests, data inconsistencies arose due to requests categorized as ‘on-demand’. These challenges, coupled with the absence of standardized collection data control, resulted in a collection characterized by (mis)organization and a sense of chaos.
Reflectively, failure to improve system infrastructure is reflected by inability to execute emerging archival processes over time. Greene argued reluctance to advance traditional archival methodologies produced commonly adopted models normalizing practices that cause greater backlogs (Greene and Meissner 2005). Simply, the normalization of ineffective processes has occurred as a response to conditions over an extended period of time resulting in reactively creating crisis management processes in place of long-term solutions (Pinto 2014).
The comprehensive study of the Collection, including its organizational management and historical context of collection processing, is crucial to establish a reliable and fully integrated corpus. This ensures the reliability, discoverability, and accessibility of the Collection’s contents. Through the examination of the Julian P. Kanter Collection as a complex case study, solutions have been developed to address the challenges associated with large-scale digital project planning, workflow management, and the standardization of methodologies. By conducting research at a micro level, it becomes possible to analyze systemic issues related to record management. Identifying and resolving these barriers that hinder the efficiency of archival information systems plays a key role in creating informed solutions and controls to prevent future archival chaos.
arrangement and original order
The initial processing of archival collections is vital for establishing cohesiveness and discoverability, as it provides a structured framework to navigate the collection effectively. Through this process, the arrangement of materials ensures the reliability of the content by defining the existence and relationships of each hierarchical component. However, when order is not maintained, it leads to inherent disorder, causing a breakdown in the archival information system. This study highlights the failures in managing the components, which resulted in a systemic collapse, disconnecting, and fragmenting the collection’s cohesion.
Upon the acquisition of the Collection by the Center in 2019, a vast amount of potential data was discovered within thousands of pages of inventories, catalogs, and notes. These documents, stored in various formats, were dispersed across multiple locations such as paper files, internal and external hard drives, servers, cloud storage, and local networks. The ‘official’ collection inventory provided to the Center in 2019 offered a partial consolidation of information regarding the Collection materials, primarily accumulated by students from the late 1980s to 2018.
The initial arrangement of the Collection in 1978 followed an alphabetical order based on the candidate’s name. It consisted of two parts: (1) an entry line containing the candidate’s name and physical format information, and (2) a brief description of the candidate, a list of related items, and, whenever possible, details about the candidate’s political party and specific election year. Commercials that did not involve candidates, such as those related to special interests, bond elections, or proposition elections, were listed separately without a specific order. In 1985, the handwritten inventory was transformed into a standardized ‘form’ that included both typewritten and handwritten text.
In 1990, a comprehensive catalog was created to document the Collection, maintaining the original order, and incorporating OCLC access numbers whenever feasible. This catalog followed the Archives and Manuscript Control format (AMC). However, it was frequently observed that the item-level descriptive data was incomplete or missing, and there was limited availability of legacy metadata.
In 2003, an updated catalog of the Collection was created, covering a total of 70,000 items categorized into three main components. The catalog provided details including (1) the last name of the candidate, (2) the election year, office, and party affiliation, and (3) the title, duration of the commercial, and a unique identification number. However, this revised catalog did not incorporate the identification numbers from the 1985, 1990, or 1996 documentation, and it lacked a coherent and logical sequence in its structure.
In 2019, the metadata of the Collection underwent significant enhancements, incorporating more comprehensive information and introducing additional elements. This update involved the retrieval of identification numbers from previous catalogs and the introduction of a new category called ‘P_COPY’.
The inclusion of the ‘P_COPY’ identification number marked a significant and fundamental change in the identification of items within the collection. Each ‘P_COPY’ group served as the highest-level container, encompassing 2 to 150 ads. Within each ‘P_COPY’ group, the ads were arranged in sequential order based on the identification numbers assigned in the 2003 catalog6.
During the initial investigation, we could not find any documentation supporting the creation of the P_COPY categorization. It is worth noting that the letter ‘P’ typically indicated that a file could be the ‘preservation’ copy. However, the process of selecting and assigning ads into the ‘P’ groupings was not identified. As we continued to (re)process the Collection, we discovered several illogical groupings within this category.
By 2022, controlled methodological workflows were developed to respond to discovered issues and increase the collection control. Legacy metadata was not discarded; instead, it was extracted and retained in collection records. Corrected or additional data was recorded using controlled vocabulary lists, providing structure to this phase of reprocessing.
The issues with system infrastructure and normalization section focuses on component identification, standardization, and other challenges that have been identified through investigative analysis and the normalization process.
Note: to review early content management techniques view – initial-codes-processes (GitHub)
Footnotes
- The Political Communication Center located at Kaufman Hall on the campus of the University of Oklahoma opened in 1983 and by 1985 held the Julian P. Kanter Archive. The Archive was purchased by the University and contained 25,000 commercials collected by Julian P. Kanter. ↩
- Alternative data from a report by The Library of Congress from 1997, “A Report on the Current State of the American Television and Video Preservation, Volume 1”, was 37,000 total commercials. ↩
- It is important to note that in 2003, there were 10,000 files which were not cataloged which is a notable factor when assessing origin of barrier creation and the history of the Collection. ↩
- It should be emphasized that ongoing efforts in working with the Collection have led to the identification of a substantial number of un-accessioned and un-inventoried advertisements spanning from 1960 to 2018. These ads, falling within the specified date ranges covered by the Collection, will be incorporated into the Collection after their official accessioning following the completion of the NSF Grant project. ↩
- As indicated in the 2012 reports and subsequent research, the reported number of Collection items varied across different sources such as reports, newsletters, and campus newspapers, ranging from 90,000 to 125,000 total items. However, as of 2019, the Collection management reported 119,000 total items. As of 2023, it is estimated that the Collection has grown to include over 123,000 accessioned items and more than 150,000 un-accessioned items. ↩
- It is important to note that documentation tracking changes to cataloging was not recovered. Sequential files varied in numbers from 2 ads to 150 ads per P_COPY folders. Files within the P_COPY folders years, offices, and candidates, or non-standard numbering. ↩
