Querying Sacred Harp’s Sonic Past through the Minutes of Sacred Harp Singings, 1945–Present
Sacred Harp Minutes is curating an expanding collection of humanities research data from the proceedings of Sacred Harp singings in a queryable database, making it possible for scholars and singers to ask a range of new questions about how participants in this music tradition historically engage its repertoire across time and space. The database is a collaboration between Emory University, the University of West Georgia, the Sacred Harp Publishing Company (SHPC), and the Sacred Harp Musical Heritage Association (SHMHA), and is directed by Jesse P. Karlsberg, Emory Center for Digital Scholarship (ECDS) senior digital scholarship strategist and vice president of SHPC. The database will initially feature more than 15,000 singings from 1945–present, including data from fifty annual volumes recently digitized by Emory’s Digitization Program and more than twenty-five born digital volumes.
Sacred Harp is a style of social hymn-singing in four-part a cappella harmony. A form of shape-note singing, the style adopts a music notation system that aids in sight-singing in which notes of different pitches feature different shapes corresponding with the syllables “fa,” “sol,” “la,” and “mi.” Singers use a tunebook titled The Sacred Harp at “singing conventions” where participants take turns leading the group in a song or two of their choice. Since the book’s first publication in 1844, secretaries have recorded and published “minutes,” listing the the names of leaders and page numbers of their chosen songs from The Sacred Harp at these singings. Published in an annual volume known colloquially as the “big minutes” since 1945, the minutes comprise a remarkably granular and comprehensive record of the decentralized lived experience of a music culture across its community of practice. Since 1995, this minutes book has been edited and published by SHMHA using a born-digital process.
The first phase of Sacred Harp Minutes involves the digitization, optical character recognition, and manual correction of the complete fifty-volume print run of the “big minutes” prior to its shift to a born-digital production process (1945–1994), drawn from four sources:
The second phase of Sacred Harp Minutes will supplement the “big minutes” with records from dozens of smaller pamphlet-bound volumes produced by state, regional, and local Sacred Harp singing conventions and associations. Our team has already documented over 150 extant minutes pamphlets from more than twenty different singing bodies dating from 1913 to 2004 in the collection of the Sacred Harp Museum, in private collections, and at university and public libraries and archives.
Please let us know if you have copies of Sacred Harp minutes pamphlets you would be willing to loan for digitization!
All fifty volumes included in the first phase of Sacred Harp Minutes and more than 100 volumes in the second phase were digitized by the Emory Libraries Digitization Program. Volumes in the collections of the University of West Georgia, the Sacred Harp Museum, and private individuals were temporarily loaned to Emory for digitization. These volumes were digitized on a Kirtas book scanner and cropped, de-skewed, and binarized by the LIMB Software.
The resulting page images are being checked for completeness by a crowd-sourced team of volunteers drawn from the contemporary Sacred Harp singing community. Page images for each volume are then uploaded to the Tranksribus platform for layout analysis and text recognition. Project volunteers then manually correct the optical character recognition (OCR) results seeking to achieve diplomatic transcript–quality accuracy. This textual data will be published as it becomes available.
If you are interested in volunteering to help correct minutes text, please get in touch!
In tandem, project team members will agree upon and transform the textual information into a unified data format for both historical minutes being digitized for this project and born-digital minutes. The database will share a data model with SHMHA’s minutes submission and editing process, enabling the ingest of minutes from future years. Volunteers will then process the data to match the agreed upon database format, publish the format to the web, and develop an API so that others can easily access the data.
“FaSoLa Minutes” is an app for iOS and Android smartphones and tablets that offers new ways to view and search the minutes of Sacred Harp singings from The Sacred Harp: 1991 Edition since 1995. Singers can use the app to identify the song that’s been stuck in your head all morning, to learn more about the favorite songs of your singing friends, to read up on a singing you hope to attend, decide on a song to lead when the one you were planning on singing has already been used, and much more. Scholars can use the “FaSoLa Minutes” app to identify shifts in the makeup of singings, popularity of songs, and participation of individuals in this music culture. The app is available in the App Store and on Google Play for $4.99 (proceeds benefit SHMHA). SHMHA hosts a more limited web-based interface for browsing Sacred Harp minutes data. Read a review of the “FaSoLa Minutes” app by Clarissa Fetrow and an explanation of its entropy number by David Brodeur and David Smead.
Every year, SHMHA publishes the Minutes and Directory of Sacred Harp Singings, a print volume compiling the proceedings at all Sacred Harp singings held during the previous year and submitted to the minutes book’s editors, along with a directory of coming singings and other information. Minutes are submitted by secretaries of singings and compiled by the editorial committee. The cost of the book is largely funded by a small fee collected from the donations of singers on the day of each singing. The minutes book includes:
The Sacred Harp Minutes project will enable more streamlined production of future annual volumes of Minutes and Directory and will also renovate the minutes submission process, creating a mobile-friendly web-based submission process that ingests data from secretaries to feed both the print minutes book and the database featured on this site.
For its first hundred and twenty years, Sacred Harp singing was confined almost exclusively to six states in the southeastern United States: Alabama, Georgia, Mississippi, Texas, Florida, and Tennessee. Beginning in the 1960s, with the support of southern singers, folk music enthusiasts in other parts of the country began forming their own conventions, modeled closely on the long-established format of historical singings. Growth in Sacred Harp outside the South since that time has been continually accelerating. Nathan Rees is using data from Sacred Harp Minutes including singing location names and corresponding GPS coordinates to map Sacred Harp singing’s recent spread.
For most of the twentieth century, minutes were published almost exclusively for singings in the six states named above; the 1985 Minutes Book documented singings in just two additional states: Kentucky and Connecticut. By 1995, there were entries for 175 singings, thirty of which took place in eighteen states outside the original six. The minutes for 2015 include a total of 289 singings. Alabama still holds the most, with ninety-two, but ninety-five singings took place in twenty-seven US states outside Sacred Harp’s pre-1970s territory. In addition, the 2015 minutes include thirty-three singings held across seven countries outside the United States.
Using Sacred Harp Minutes data to systematically identify songs from The Sacred Harp represented in a corpus of recordings featuring 3,890 songs at Sacred Harp singings, this project uses a custom-designed beat tracking system to analyze how tempos selected by practitioners of this decentralized music culture correspond to historical prescriptive instructions for appropriate tempos. Singers at Sacred Harp singings take turns leading songs and have discretion over tempo. The 1844 Sacred Harp prescribed tempos in seconds per measure for each of the book’s seven “moods of time,” though as music educator Allen Britton noted in 1949, “whether or not the exact tempos ascribed to the various signs was strictly observed in practice we cannot tell.” A 1911 revision to the book’s introduction removed these instructions.
This project assess conventional wisdom among singers, which holds that tempos have increased during the twentieth century, determining that tempos for some moods of time have indeed increased, but that observed tempos for a majority are remarkably close to original prescriptions. The project suggests new approaches to extracting tempo information from beat onset data and structuring corpora used to evaluate beat tracking systems.
The digitized minutes also make it possible to associate dozens of previously inaccessible private collections of home-made Sacred Harp singing recordings with time and place by documenting a sequence of songs in a given recording and algorithmically matching the sequence to minutes data. Numerous Sacred Harp singers from Georgia, Alabama, Mississippi, Florida, and Texas have collections of recordings they or their singing friends and relatives made at Sacred Harp singings since the 1950s on transcription discs, open-reel tapes, audiocassette tapes, or in born-digital formats. Few of these recordings have been digitized and many lack identifying information. The Sacred Harp Museum of SHPC and Pitts Theology Library at Emory University have partnered with Osiris Studio in Los Angeles to offer at-cost digitization and archiving of these recordings with the goal of preserving Sacred Harp recordings and increasing the breadth and accessibility of a corpus of known Sacred Harp recordings for additional research.
Even a sequence of a modest subset of songs sung at a given singing comprises a unique fingerprint of the event. The project uses a custom=designed algorithm to match these sequences of songs identified in the recordings with information from the Sacred Harp Minutes database to determine likely matches for these unidentified recordings. In the future, the team hopes to use identified singings as training data to develop a computational listening program to automatically identify sequences of songs in recordings to identify singing recordings without a need for manually documenting the content of the recordings.
This project’s hosting and some other infrastructural costs are funded by two non-profit organizations, SHPC and SHMHA. The University of West Georgia’s Annie Belle Weaver Special Collections at the Ingram Library and the Sacred Harp Museum at SHPC loaned volumes from their collections to Emory for digitization. Emory University’s Pitts Theology Library, ECDS and Emory Libraries contributed technical expertise, digitized pre-1995 minutes volumes, and supported the project’s file-sharing infrastructure.
Laura Akerman, Melanie Albrecht, Mairye Bates, Kevin Beirne, John Berendzen, Adam Berey, Stacey Berkheimer, Justin Bowen, Marie Brandis, Morgan Bunch, Steve Cackley, Judy Caudle, Leigh Cooper, Kate Coxon, Emily Crespo, Clarisa Fetrwo, Ann Riley Gray, Carol Huang, Sarah Huckaby, Stephen Hutcheson, Robert Kelley, Nancy Mandel, Dorothea Maynard, Marian Mitchell, Angela Myers, Michael Ruhl, Dawn Stanford, Melissa Stephenson, Mary Amelia Taylor, Tivey, Judy Van Duzer, and Micah John Walter.