Using Database Development to Facilitate Undergraduate Research in the Humanities
Editor’s Note: This is the first in a series of blog posts about undergraduate research.
Esther M. Morgan-Ellis is an associate professor of music history.
Humanities faculty often find it difficult to involve undergraduates in our own research. Most of us are used to working alone, and we are usually engaged in projects that do not lend themselves to collaboration. For this reason, humanities undergraduates often find themselves at a disadvantage in comparison with their peers in the sciences. All the same, these students will benefit from opportunities to pursue original research in collaboration with faculty, and there are ways we can get undergraduates to become meaningfully involved in our research projects. Here, I will describe how developing a database to inform my own research on the community singing movement allowed me to collaborate with students to our mutual benefit.
The Database of American Sing-Along Repertoire (DASAR) tracks the appearance of individual songs across songbooks and sing-along media. As of this writing, it includes 1,756 unique song titles and 4,674 correlations between songs titles and media objects, which include sing-along LPs, short films, and radio and television broadcasts. DASAR, however, is very much a work in progress, and new songbooks and other media are documented on a regular basis. Through DASAR, I publish the raw data that underpins many of my arguments and conclusions. Other scholars are welcome to examine this data for themselves and to use it in any way they see fit.
I began to develop the database that became DASAR in 2018 as part of my research into the radio program Gillette Original Community Sing (CBS, 1936-37). My archival research had generated a complete list of songs that were associated with the program, which created an opportunity to examine trends in the repertoire. After fumbling with Excel spreadsheets, I got assistance in creating a MySQL database and entered every selection that was sung on the air or included in one of the two accompanying songbooks (the repertoire for each broadcast can be browsed here, while the repertoire included in the two songbooks can be viewed here and here). Now I was able to answer questions regarding the frequency with which certain songs were sung, the relative popularity of songs from certain eras (e.g. Civil War songs, Gilded Age songs), and the relative prominence of various composers and lyricists in the sing-along repertoire. I also tagged every song with a topic, which allowed me to track what Gillette participants were singing about. However, I began to wonder how this repertoire compared to others of the era. To find out, I entered the contents of prominent songbooks and sing-along short films that were exhibited in the 1920s and 1930s. It quickly became evident that I had developed a powerful tool for understanding how various sing-along repertoires interacted with one another and changed over time.
In Fall 2018, I secured a mini-grant from the UNG Center for Undergraduate Research and Creative Activities to fund the involvement of college students in the development and exploitation of DASAR. I was able to hire four students—sociology major Devin Hing and music education majors Ashlynn Nash, Emily Nelson, and Anita Ingram—to work with the database for 100 hours each. I assigned each student a stack of songbooks and trained them in data entry. I also talked to them about my own work and shared the questions that I hoped to answer using the database. After a couple months logging songs, each student developed a research question and used the database to frame or resolve that question. Hing entered all of the songs mentioned in reviews of sing-along organ solos presented in picture theaters over a three-month period in 1927 and analyzed the results, discovering in the process that half of the theatrical repertoire was published by a single firm: Leo Feist, Inc. This significant research outcome has transformed how I think about the practice of picture-theater sings, which were clearly shaped by the interests of publishers to a remarkable extent. Nash searched newspaper databases for evidence of African American sing-along practices and repertoire, while Nelson interrogated the underrepresentation of female composers in DASAR. Ingram, who had documented songbooks produced on behalf of the Music Supervisors’ National Conference, wrote about the history of community singing in American music classrooms. All four projects increased my own knowledge of the community singing movement and gave me new ideas for future research. They also benefited the students, all of whom presented their work at the 2019 UNG Annual Research Conference and two of whom went on to complete additional independent research projects under my supervision.
In early 2020, I decided that DASAR had become substantial enough to merit publication. I purchased the domain dasar.us and arranged for the database to go live. The online version of DASAR does not provide full functionality. While it includes comprehensive search tools, complex queries still need to be entered into the MySQL workbench itself. Similarly, I have not yet included the “Events” category, which documents the use of songs in the context of specific historical events, nor have I included the subject-matter tags. Neither of these features has been implemented to the extent necessary to become useful. My reasons for streamlining the online version of DASAR had to do with feasibility and usability. On the one hand, programming the many queries I use into the website would be prohibitively difficult. On the other, I want DASAR to be user-friendly, and to invite browsing by anyone who might be interested in songbooks or sing-along media. Scholars who want to interrogate the data are welcome to contact me with their queries.
This last point speaks to my dearest hope for DASAR, which is that it will foster collaboration. In additional to enabling the involvement of undergraduates in my research, this project has led me to make connections with scholars and practitioners who share my interests and who are eager to contribute. I have put together a spreadsheet template that allows collaborators to easily document and submit songbooks or other media for inclusion in the database. In return, I can perform analysis on behalf of contributors, who gain new insights into their own data. Currently, DASAR reflects my own research interests, but I hope that soon it will reflect the interests of a scholarly collective. At the same time, I hope that other teachers will take the opportunity to involve their students in data entry and analysis. Database projects accommodate participation at various levels. Sophisticated researchers can take full advantage of analytical tools, while even the novice can contribute data or run simple queries.
Nobody benefits from DASAR more than I do. I query the database frequently in my research and have used it to inform my conclusions about historical sing-along practices. Most recently, I documented all of the audiovisual sing-along media released by Mitch Miller in the 1950s and 1960s. This included all of the participatory selections featured in the television program Sing Along with Mitch (NBC, 1961-64) and the songs included on his many sing-along albums. A collaborator who I met through this project is currently documenting her collection of Mitch Miller songbooks. Some of this information—the contents of the albums, for example—is freely available online, but the list of songs used in the television show is preserved in the Mitch Miller papers at the New York Public Library and can only be viewed by visiting the collection. My ability to publish this information makes it available to my new collaborator, who in turn is able to grant access to materials that reside in her private collection. We both benefit from the mutual exchange of ideas and knowledge, and anyone else pursuing a related line of inquiry benefits from the fact that our data is publicly available. It’s a win-win-win.
Students are able to enter this scholarly exchange on their own terms. DASAR can be used purely as a reference; for example, a student writing about a particular song can use the database to track down songbooks or media in which it appears. Because DASAR is a collaborative and open resource, however, students can easily become engaged in its development. It takes minimal knowledge or ability to make a meaningful contribution. As a result, a project like DASAR encourages dynamic pedagogy in which students collaborate with instructors to shape scholarly discourse and become the creators of knowledge.