The study of media and film has long relied on historical objects and media archives. While the advent of new digital technologies has brought a flood of new research focusing on the present (e.g. social media, wearables, smart cities, etc.), we argue that archivally-based media history projects are equally valuable for helping students re-examine accepted histories and media objects, as well as uncover alternative and often silenced histories that accompany any new medium.

This article reports on two related media history projects that we lead at the University of Wisconsin-Madison: the Media History Digital Library (MHDL, https://mediahistoryproject.org) and the PodcastRE database (http://podcastre.org). Each project presents pedagogical opportunities by offering researchers vibrant collections of historical and contemporary media in a highly digital, searchable and re-searchable format. The MHDL is a free, online resource dedicated to digitizing historic books and magazines about film, broadcasting, and recorded sound for broad public access. The collection contains millions of pages of scanned texts, as well as analytical tools to search and explore trends in media history. PodcastRE is a free and searchable database of podcasting culture. The site indexes and preserves audio and metadata on over 4 million podcasts from the last 15 years and makes audio resources easier to search, analyze and visualize for researchers and teachers. While the former project uses historical pre-digital documents to ask new questions, the latter brings historical questions to bear on born-digital formats. By encouraging students to engage critically with media history, as it occurred and as it is currently unfolding, we hope to raise questions about 1) how archives are built, organized and displayed, 2) which media artifacts are saved or forgotten and why, and 3) how metadata and object descriptions shape and influence discovery and interpretation of historical objects. The benefits of these learning outcomes extend beyond the classroom – helping students understand the role preservation plays in their own lives, as both collectors and creators of media artifacts.

We approach these projects from the perspective of media history and communications studies scholars first, and only secondly as “accidental archivists”. While both of us have become familiar with the challenges and potentials of digital (and physical) databases through our projects, neither of us is an archivist or information scholar by training. Eric Hoyt’s new role as Director of the Wisconsin Center for Film and Theater Research has brought preservation questions to the forefront of his recent work. Jeremy Morris’ early involvement in podcasting communities in the mid-2000s has helped fuel his interest in preserving the many voices the format encourages.

The Media History Digital Library: Training Students to Think Critically about History and Technology

The Media History Digital Library’s search engine, Lantern (http://lantern.mediahist.org), enables students of film and broadcasting history to run queries across a 3 million page corpus of books and magazines. Led by Eric Hoyt and the Wisconsin Center for Film and Theater Research, the MHDL is freely available to users worldwide and dedicated to broad public access. The resource exists because of ongoing collaborations, facilitated by the MHDL, among institutions and collectors (which lend materials for scanning) and sponsors (who pay for the scanning). When a researcher downloads a page of Broadcasting or Moving Picture World, for example, they are the end-user of a digitization process that began with taking the magazine off the shelf of a private or institutional collection (for instance, the Library of Congress or Museum of Modern Art Library) and touched many hands, computers, and algorithms before becoming searchable data.

We believe that it’s important to highlight the collaborative nature of the digitization process for two reasons. First, it represents an open access, non-profit model for the building of digital resources that is far more sustainable for university libraries than paying for yearly subscriptions to commercial database services. We hope more digital projects and universities will move in this direction. Second, we believe that incorporating an understanding of provenance and digital transformation into our teaching is important for training students to think critically about both media history and the digital world in which they live.

One assignment that can be effective for accomplishing this goal is an artifact curation project. When we used this in our Cultural History of American Broadcasting course, we asked students to select one digital artifact from the MHDL or a local archive, describe it, analyze it, and argue for its significance to radio history (click here to see the full assignment handout, which JCMS readers are welcome to revise and use; an additional dozen assignments using the MHDL’s collections in film and media courses can be found in the MHDL’s Resources for Educators section). Among the many benefits of this assignment is that students actively perform the role of media historian — selecting and interpreting primary sources, drawing meaning — yet it is on a scale that is manageable within the context of their busy lives. Although many students wanted to immediately jump to the analysis and argument stages, we found that the description stage actually challenged them the most and proved to be essential for the assignment. Students had to answer a series of questions: Who created the artifact? When, where, and why? Who was the original audience? Why was it saved and by what process did it become accessible to you? What is the provenance? By addressing these questions, students developed more awareness of why it is they have access to some sources and not others.

For deepening this understanding of presence and absence in digital collections, we recommend an in-class exercise that employs the pedagogical strategy known as POE (Predict, Observe, Explain) with the MHDL’s data visualization app, Arclight. We have described this exercise at some length on the Project Arclight blog. But the central idea is that students guess what the big data approach might yield, observe the results, compare the results to their hypotheses, and finally dig deeper to understand why the Arclight app returned the visualization that it did. The answers sometimes tie back to historical phenomena of central concern to the class (for example, the term “football” appears prominently in the MHDL’s collection during the mid-1930s due to the popularity of college football radio broadcasts and short films and their frequent mentions and advertisements in entertainment trade papers). But the answers can also lead back to critical reflections on the underlying collections and technologies. Due largely to intellectual property restrictions, the MHDL’s collection primarily centers upon US publications created prior to 1964. A search for “Michael J. Fox” won’t return anything about the 1980s screen star because of these limitations of the collection. The technology fails at times, too, especially when it comes to OCR (optical character recognition) generated from publications that featured unusual fonts and typefaces. These meta-reflections about archival absence and digital technology can be some of the most valuable learning outcomes from this exercise. And they are equally applicable how students engage with collections of born-digital objects, including PodcastRE.

The PodcastRE Database: Archiving and Surfacing New Voices

Since the early 2000s, there has been a steady and impressive growth in podcasting. While there are still significant disparities in terms of which voices and bodies get to host, produce, and find success through podcasting, the format’s emergence has led to a flood of new voices, sounds and perspectives, and has created a treasure trove of stories, conversations and ideas that can help us teach alternative histories that offer fuller and more nuanced pictures of the past (and present).

We built the PodcastRE.org database to track and preserve this burgeoning sonic boom, but also as a teaching tool to help instructors looking to incorporate a broader range of podcasts in their course syllabi. Part of teaching media history and media archives involves showing students that the ways a particular database or archive presents and makes available certain resources and not others affects the kinds of resources that get used and studied and the kinds of histories that get written. Finding the commercial podcasting outlets (like Spotify and Apple) lacking in this reflexivity, we worked consciously to create search tools that would prioritize a breadth of discovery opportunities, rather than returning the most popular, recent or dominant search results. For teachers and students, the site provides a valuable resource for searching through podcasts by keywords, categories, date, and other metadata to return results that are not solely based on models of industry success, popularity or personalized recommendations (as on Spotify or Apple). We also provide visualization tools for users to graph keywords over time to see, for instance, how frequently a term like “Me Too” appears in our database over a particular range of years, or to create word clouds based on keyword metadata. By prioritizing connections between keywords, rather than algorithms that surface the most popular or most downloaded shows, we hope to provide access to a wider range of podcasts than other similar collections.

The PodcastRE database also serves as a conversation starter for students around questions that face all archives and databases: which artifacts do we save and why? how do we store and make visible those artifacts? We believe that teaching with and about archives also means teaching why some objects are saved and not others, why some objects surface while others stay buried. Although the digital nature of podcasts means we can gather a huge number of podcasts in an automated fashion, we wanted to balance what was possible with what was necessary for building a database that reflected podcasting’s full breadth. Our site ingests the most popular shows (i.e. by automatically collecting the top 100 podcasts in Apple Podcasts in several different regions), since we wanted to track which shows were being framed as popular in the podcasting ecosystem. But we have also added specific podcasts and content from podcast networks that represent sounds and sonic perspectives from particular communities and identities that are often less audible in other media or other podcast aggregator sites. Thanks to content cues from existing collections like Podcasts In Color or Women in Podcasting, as well as podcast networks like Indian and Cowboy or Potluck: An Asian American Podcast Collective, we try to go beyond the Apple Podcasts charts. We work with scholars researching under-represented podcast producers/content for recommendations and we offer a “submit a podcast” public feature so any podcaster, student or user of the site can include a podcast in our collection. Even with these intentional efforts, the site is still heavily English language-based, which introduces further blind spots; we continue to build features and mechanisms to bring in a more diverse range of content and voices.

We hope this collection will allow for class assignments that think across the variety of podcasts in the collection. Instructors could, for example, design an assignment where students are asked to use PodcastRE to study a particular media controversy, like Colin Kapernick’s decision to kneel during the anthem at NFL games. Students could search the database for terms like Kapernick or more specific hashtags or phrases (e.g. #LetKapPlay) and compare the ways in which different podcasts (i.e. Industrial/professional podcasts like ESPN’s 30 for 30 or The Glenn Beck Radio Hour or more amateur and hobbyist podcasts like The Follow Through with Clipps and Drew) provide different means of assessing, understanding and discussing the story, especially if the podcasts are hosted by communities most directly affected by the controversy. Using the term frequency line graph, students could also see the number of podcasts in the database talking about the topic over specific periods of time. Using the associated keyword cloud visualization, they could also analyze what keywords appear alongside “Kapernick” across all podcasts in the database using that keyword (e.g. both Muhammad Ali and Donald Trump appear in the Kapernick word cloud, encouraging students to think contextually and historically about media events). Given that this keyword metadata is often created and chosen by the podcasters themselves, it offers a form of self-representation and self-classification that is worth further analysis and provides insights into the individuals and communities making podcasts (for more see Hoyt et. al. 2021 or Susan Noh’s chapter in Saving New Sounds: Podcast Preservation and Historiography).[1]

Ultimately, both PodcastRE.org and the Media History Digital Library are tools meant to teach students the value of archivally-based media research, past and present. As these students go on to create digital media of their own - either as students submitting podcasts, video essays and other multimedia assignments or as professionals making the next wave of news and entertainment content - we hope the lessons from PodcastRE and the MHDL prompt students to think of the value of media artifacts as both tools for doing history and artifacts of history. More importantly, we hope they learn the importance of preserving the stories and voices that media objects generate.

Author Biographies

Eric Hoyt is the Kahl Family Professor of Media Production in the Department of Communication Arts at the University of Wisconsin-Madison. He is the Director of the Media History Digital Library and the Wisconsin Center for Film and Theater Research. He is the author of Hollywood Vault: Film Libraries before Home Video and Ink-Stained Hollywood: The Triumph of American Cinema’s Trade Press, as well as co-editor of The Arclight Guidebook to Media History and the Digital Humanities, Hollywood and the Law, and Saving New Sounds: Podcast Preservation and Historiography.

Jeremy Wade Morris is an Associate Professor of Media and Cultural Studies in the Department of Communication Arts at the University of Wisconsin-Madison. His research interests include the digitization of cultural goods and commodities, the history of software, and the current state of the popular music industries. He is author of Selling Digital Music, Formatting Culture (2015), co-editor with (with Eric Hoyt) of Saving New Sounds: Podcast Preservation and Historiography (2021) and co-editor (with Sarah Murray) of Appified: Culture in the Age of Apps (2018). He is also the co-founder of PodcastRE.org, a searchable, researchable database of over 4 million podcasts.

Works Cited

• Hoyt, Eric, Bersch, JJ, Noh, Susan, Hansen, Sam, Mertens, Jacob and Morris, Jeremy (2021). “PodcastRE Analytics: Using RSS to Study the Cultures and Norms of Podcasting”. Digital Humanities Quarterly.

1. [Noh, Susan (2021). “A RE-emphasis on Context: Preserving and Analyzing Podcast Metadata”. In Saving New Sounds: Podcast Preservation and Historiography. Jeremy Wade Morris and Eric Hoyt, eds. Ann Arbor: University of Michigan Press.]