Digital Sites of Binder's Volumes

Many digital tools, platforms, collections, and websites have been developed that assist humanities scholars with their inquiries. These can be used as substitutes for handling a particular artefact, as tools to assess data across a vast number of items, or for many other reasons. The discoverability of information should constitute a crucial purpose in the creation of websites for which a main purpose is sharing facts that cannot be easily gotten by other means. Thus, if one wishes to find statistical data on a given publisher, printer, or book seller, their names must be identified and tagged as searchable. Ideally, GPS (global positioning system) identifiers further enhance the serviceability of such sites. In the case of sheet music, the names of composers, titles, and details of language, dedicatees, derivative works, and similar information tend to be the means by which cataloguers, in the very least, have marked records so that users may locate specific pieces and editions. Many institutions have crafted websites where it is possible to search for sheet music in this way; the Sheet Music Consortium (https://digital.library.ucla.edu/sheetmusic/) is the most comprehensive site in this regard.¹

But none of this evidence reveals who used the music and how. For this, the best artefacts are bound volumes of sheet music that include the owner's name, or the names of family members, friends, teachers, and other marginalia scribbled anywhere on the pages sewn together and fortified with strong front and back boards (and spine), and often covered, in part at least, in leather. For the nineteenth-century, a substantial number of these have the owner's name embossed on the front cover – often the first clue to the item's history – and an overwhelming number of them belonged to women. When viewed as a whole, these collections illuminate the lives and experiences of underrepresented historical actors who constituted a major portion of the population but whose lives have not been deemed worthy of study. Furthermore, they have not been preserved with as equally purposeful documentation as more familiar research resources. Regularly cast into boxes buried deep within archives, these materials have been deemed non-essential research resources, ephemera, yet they reveal how musicking happened during the nineteenth century (the heyday of this type of source).²

Several scholars, including as Jeanice Brooks and me, have interrogated these collections for what they reveal about musical culture in an explicit time and place as well as broader implications discerned from considering many such volumes.³ What began as isolated studies conducted in a few places has now grown markedly to encompass almost any place where western music was heard. In recent years, this work has challenged generalizations across music history narratives, a fact that would seem to suggest that archivists and those in related fields should make them readily available for research. The question of digital or analogue resources remains immature at this point because hardly any archive, library, or other institution has crossed the divide and approached their holdings from the perspective of digital alone – but more on this later.

At a basic level, a key hindrance to this process is the lack of a common term to describe these volumes – if the items are designated at all. Some prefer ‘bound volumes of sheet music’, but others find this cumbersome. With the publication of Emily's Songbook in 2011, US-American musicologists adopted the term ‘binder's volume’, and although this term has become standardized by those working with such sources in the United States, the adoption of ‘binder's volume’ as a searchable entity (in, for example, a keyword search) is far from universal, even in US-American libraries and archives.⁴ Moving to an international level, the range of descriptors grows exponentially. In Chile, for example, scholars use the term ‘álbum’, but this connotation carries differing meanings in Germany. Some Viennese researchers have adopted ‘binder's volumes’ when writing about British and US-American collections, but not all – and certainly not Viennese archivists. To find these types of items in the Bibliothèque nationale de France, for example, one must incorporate a variety of search terms and processes to locate the items, and this constitutes the most common way to find them. All of this ambiguity reflects the newness of binder's volume research and the unfortunate consequences of not being able to ask for an item when visiting an archive.

Myriad searches can be employed to find whether an archive houses binder's volumes. One of the simplest involves making an educated guess of a popular composer or title for a given period. For example, ‘Home! Sweet Home’ (with or without the exclamation point), ‘Casta diva’ or ‘Norma’ (from Bellini's Norma) or ‘When the Swallows Homeward Fly’ (Franz Abt) work well for US-American searches of the mid-nineteenth century. Granted, these may yield many search results, as they were the most popular melodies of the period, but there are ways to determine if they exist in binder's volumes. The most common cataloguing method that I have encountered gives the individual title's position within a call number. For example, a search in the Bodleian Library for the title ‘Una voce poco fa’ (Rossini) revealed a score with the call number Mus. 302 c.15 (8) (belonging to Augusta Norton). The way it appears shows that the piece I searched is the eighth one in a volume, which suggests that it is a binder's volume. Once an archive's cataloguing method has been discovered, hopefully other clues direct further searching there.

Despite the difficulty in locating binder's volumes, they have proven ripe for digitization projects.⁵ Some of the most informative websites stem from such projects and exist independently (to a degree) from archives and libraries. Although they may be housed under such an institution, the funding for the research required to produce the site came from outside the institution, and these must be seen differently from those attached to an archive.⁶ Two impressive sites illustrate this approach. Funded largely by the National Endowment for the Humanities, American Vernacular Manuscripts, ca. 1730–1910 (https://w1.mtsu.edu/popmusic/AmericanVernacularMusicManuscripts.php) allows the viewer to search a database of over 10,000 manuscripts by keyword or song/tune title.⁷ High quality images of each page result from these searches, and links to detailed descriptions of the manuscripts accompany each source. Moreover, another link allows the viewer to explore the entire manuscript collection as a whole entity – a crucial need for the type of research described above. People at two institutions housing significant collections (Middle Tennessee State University and the American Antiquarian Society) collaborated for this project. It forms a digital collection, meaning a collection that only exists digitally. Directed by Dale Cockrell and Thomas Knowles, the project's website launched in 2014.

Following an entirely different approach, Jeanice Brooks led a project that made available digitized images of manuscripts associated with Jane Austen (https://archive.org/details/austenfamilymusicbooks, last updated in January 2022). Each bound volume presents as a distinct item that includes a description of the manuscript itself, provenance, location, and other details. The viewer can search metadata and text, making the site an efficient tool for research on the period, specific pieces or composers, and many related topics. Housed at the University of Southampton, the collection of materials covered extends well beyond that institution. It, too, is a digital collection.

To embark upon a project like these two, a researcher must first have found the current location of materials whose significance has been previously identified or (as often happens) discover them by accident while researching something else. This task is not easy. Most people turn first to libraries and archives, a situation that stresses the usefulness of online catalogues. The amount of information about individual binder's volumes that is readily available online differs tremendously from institution to institution. If MARC (MAchine-Readable Cataloging) records (or similar behind-the-scenes pages) can be accessed, the viewer may begin to ascertain whether the binder's volume figures in their research project. Because these are notoriously inconsistent in details, however, one should not count out a resource if the record lacks substantive data. Several factors influence whether an archive has populated many fields for binder's volumes, and often they are out of the hands of the music staff (if one is available). For example, administrative powers-that-be must designate funds to employ staff to spend the time necessary to look at each individual item and enter the necessary information. Moreover, some institutions do not retain music specialists who can discern what is necessary, resulting in haphazard documentation. With this in mind, the details of some large collections remain undiscoverable – but not at the fault of the staff.

Perhaps the most glaring example in this case is the Library of Congress (LOC), which houses over 300 binder's volumes, none of which is easily discoverable. An LOC call number for these items has been developed (M1 .A15 vol. #), and this proves the simplest introduction to their holdings: it is possible to enter this call number (and a specific volume number) in order to retrieve the entire collection.⁸ When the decision was made to catalogue these items (years ago), the cataloguers followed a system that had been crafted even earlier, which is to list each individual piece by composer and title, but not the volume as a whole.⁹ To date, budgeting has not provided for much further work on this vast collection, although a few scholars have worked on producing more extensive data in their own research.¹⁰

The amount of information made available to the viewer proves crucial in accessing binder's volumes online. Here I do not differentiate between born digital and analogue sites – if I did, this review would be extremely brief. Rather, I focus on whether a site indicates such factors as timeframe of publication, locale of use, age of the compiler/owner – as much information as the person entering the data has time to include. The amount of data and how it is presented determines the usability of websites, and a wide range of options can be found throughout libraries around the world. This is, afterall, digital information. As of spring 2024, the most extensive websites for researching binder's volumes exist at the Houghton Library, Harvard University, and in the Music Library at the University of North Texas (UNT). Both exemplify the range of data a researcher can discover and use without ever visiting the physical site.

While working with Brooks at the Houghton Library, Andrea Cawelti (Ward Music Cataloger, Harvard Faculty of Arts and Sciences) realized the value in binder's volumes as resources. She collaborated with Robert Cunningham (Chief Music Cataloger, Boston Public Library), and together they devised a scheme to parse data in MARC records that should influence future cataloguing procedures. Cawelti has been a leading advocate for the importance of binder's volume research in the Music Library Association (MLA) and continues to assist researchers who are interested in these items at the Houghton Library. The information in Harvard's search engine (Hollis) is analogue, and the volumes belong to the larger collection. (Most are part of the Theatre Collection.) Nevertheless, one can explore individual volumes with relative ease because of the depth of information provided on the website. Cawelti adopted the term ‘bound music collection’ to name these objects, and that is the easiest way to locate them in the Houghton collection.

For the researcher exploring ideas such as social relationships, a distinct advantage of the Houghton site is the biographical information available for certain items. Not every binder's volume's landing page includes this extra data, but those that do remove a step from the viewer's workload. Funding levels have not allowed staff to delve deeply into each of the 360 binder's volumes in the Houghton Library, but it proves a significant help when available. Gleaning information from such resources as ancestry.com, Cawelti includes helpful data from a compiler's biography that will assist future research connected to her binder's volume and the myriad directions such scholarly work may go. The entirety of the Houghton's holdings range beyond both ends of the nineteenth century, and many have a provenance of the northeastern United States. For example, the description of volume 14 of Catharine Flint's collection (https://id.lib.harvard.edu/alma/990147476270203941/catalog) contains a wealth of information about this binder's volume and the compiler herself (see Fig. 1). Other volumes associated with her are listed, and non-musical contents (a calling card in this case) are noted.

Fig. 1 Catharine Flint vol. 14, Houghton Library, Harvard, beginning of the catalogue record

Like all others catalogued in Hollis, the record for Flint volume 14 lists the number of pieces; their titles, composers, and publication information; the year of compilation; subject headings; and other relevant data a researcher might find useful in developing a topic related to her collection. This is evident from simply looking at the contents list. Flint chose to open this binder's volume with a piano arrangement of Beethoven's sixth symphony – an unusual item for a woman's collection before 1860. That she intentionally did so speaks to Beethoven reception in the United States, her own shifts in repertoire preferences over time, and other routes of investigation.¹¹ She also included three of Ferdinand Beyer's fantasies based on Italian operas, which are more representative of popular tastes in the 1840s and 1850s. In addition to the objective data presented for each volume, the notes offer the viewer a general sense of each binder's volume, with details such as pressed flowers/leaves, indices, and similar data typically absent in MARC records.¹² This particular record includes some digital images of pages within the volume and the front cover (with nameplate), although not all of the Houghton volumes have been so lavishly recorded (see Fig. 2).

Fig. 2 Sample pages from Catharine Flint vol. 14, Houghton Library, Harvard

All of the contents appear in a single paragraph because the MARC data has been listed under a single line entry, and some viewers may find such a presentation hampers their ability to see the listing as clearly as on other sites (see Fig. 3). This situation harkens back to the decision to list by compiler or by individual title within a volume described above. For rare or unusual content, Cawelti additionally catalogued individual pieces. Nevertheless, although a search of the title can be found on a given binder's volume's landing page, titles are also searchable in Hollis. In other words, if one tried to find ‘The Merry Maid Waltz’ in a general catalogue search, this volume would be one of the results.

Fig. 3 Contents of Catharine Flint vol. 14, Houghton Library, Harvard

At the University of North Texas, Maristella Feustle has taken an alternative approach to that of Cawelti, one that envisioned a digital starting point. Importantly, although Feustle and Cawelti envisioned some differences in how MARC records would be allocated, much of the same type of information can be located on each site. In this, the type of data recorded (composers, titles, size, etc.) overlaps between the two women's work. Feustle's expertise with digital platforms and their capabilities influenced her work in this direction. She began working on binder's volumes in the UNT collection after Cawelti had established her procedures at the Houghton Library, and she has had the luxury of being able to include an image of each page of the collection's binder's volumes on the website. (Granted, the number of volumes she works with, about thirty, is significantly smaller than the Houghton's collection of just under four hundred.) Whereas Cawelti included some of these images for identified volumes, Feustle incorporated each page since she organized hers as a visual presentation from the beginning.

The binder's volumes are housed under the UNT Digital Library. They can be found directly at https://digital.library.unt.edu/explore/collections/UNTML/browse/?fq=dc_type:image_score, although they can be located without this link, which someone just beginning their research may not have. A more typical approach is to start from scratch. A search of ‘binder's volume’ on the main page (https://digital.library.unt.edu) yields almost 2,000 results, but whittling it down by selecting ‘Musical score/Notation’ under ‘Resource Types’ brings up the thirty volumes in their collection. These vary according to dates of compilation, but as a whole the group represents women's collecting practices across the nineteenth century.¹³ Once on a given volume's landing page, the ‘Overview/Who/What/When/’ links in the top left take the viewer directly to that information if desired (Figs. 4 and 5). Each individual title within a binder's volume has been catalogued, thus it is possible to search by title, composer, publisher, as well as conductors, performers, or dedicatee (if applicable). Lyricists appear under ‘Other’. Every landing page has a list of each of these (with brief explanations of who they were), the size of the volume, an image of the front cover, and options to view each page. Perhaps the most innovative choice here is a link to view the Metadata API (Application Programming Interface) in several different formats: Dublin Core (Text, XML, JSON, and RDF/XML), UNTL (XML), and Access METS (XML). This capacity allows the viewer to capture data and incorporate it into several different ontologies for their own use.

Fig. 4 Page from binder's volume associated with Mrs. Phillips’ with a signature contemporary with the music therein¹⁴

Fig. 5 Page from binder's volume associated with Mrs. Phillips’ showing a later handwriting style

The volume with ‘Mrs. Phillips’ embossed on the front exemplifies the ways to use this site for research. For example, the first piece, ‘Dussek's Celebrated Concerto with the Plough Boy’, has ‘Miss L. Henley’ written in pencil across the top (see Fig. 4). To find out if others have the same name, Feustle has made it possible to view thumbnails of every page. One other page has ‘Miss Henley’ written on it, and another has solfege indicated above the vocal line. This prompts the question of Henley being Phillips before marriage, however, the style of handwriting of Mrs. J. M. Phillips on p. 3 seems to be a later one and not contemporary with the music bound inside (see Fig. 5). These items are easily detected from the thumbnails of all the pages (a feature of this site), making skimming the volume for clues about its use relatively simple. From these it is possible to begin to tell the story of this volume.

These sorts of details demonstrate the functionality of the UNT website: Many aspects of the physical source can be ascertained without having to travel to Texas. The landing page for each volume provides as much information as the person cataloguing it can glean in a quick perusal, and further information can (and probably must) be found in other locations (digital and physical).¹⁵ Figure 6 illustrates that the range of data extends not only to composers but to others mentioned on the printed page and detailed facts about the publisher (and other pieces in the volume from the same publisher). All names that appear in multiple places are linked to others in the UNT digital library.

Fig. 6 Screenshot of landing page for ‘Mrs. Phillips’ binder's volume (https://digital.library.unt.edu/ark:/67531/metadc1706054/)

The digital presentation of UNT's binder's volumes illustrates how significant these objects are as items themselves. This difference immediately strikes the viewer upon opening one of the binder's volumes landing pages or by choosing the option to view all of the pages (see Fig. 7). It also goes far beyond that of most sites in that it includes dedicatees, names mentioned otherwise in its pages, and other types of information not found on most websites. Indeed, Feustle's work at UNT should constitute the model for how these objects should be presented in the future – if funding allows, of course. If not, Cawelti's analogue example shines as the most useful for discoverability when dealing with binder's volumes, because one can search composer and/or title (if that be the emphasis of a research project) or ‘bound music volume’ (if the materiality of the volume is the focus).

Fig. 7 Screenshot of Gertrude K. Whitaker's binder's volume pages

And here, this review takes an unexpected turn in that I introduce my own work into this discussion. In 2023, Feustle, Brian Anderson and I were awarded a Humanities Collections and Reference Resources Foundation grant by the National Endowment for the Humanities for a project, entitled the Binder's Volume Research Initiative (BVRI), that addresses the needs and concerns of those researching binder's volumes. We have isolated two main goals for this stage of the work. The first is to establish consistent guidelines for cataloguing these objects and to present them digitally in an easily discoverable format. Working with over ninety columns of identified data items, we seek to provide the viewer with an array of information stretching from names of friends and family members to GPS data. We link to resources outside our own creation, such as the Social Networks and Archival Context (SNAC) Cooperative (https://snaccooperative.org), as well as across our own digital collection (the BVRI) so that one can discover relationships (schoolmates, family, teacher–student connections) not necessarily apparent from investigating a single binder's volume.

The second goal of the BVRI is to find a way for sites constructed on varying platforms (pre-existing sites, analogue or digital or with dissimilar MARC data parsing) to be discoverable and coordinated. In other words, how do we make an older site, such as that at the University of North Carolina (https://dc.lib.unc.edu/cdm/landingpage/collection/sheetmusic), interface with a newer one, such as that at the Houghton Library? This second stage informs how user-friendly our site will be in the future. Very much in the beta phase of development, the project has completed its first year and holds the promise of vast potential in research on the musical culture that accompanied the heyday of binder's volumes (as well as the birth and decline of this practice).¹⁶ We have tested the waters at the 2024 national meeting of the Music Library Association, where the current version of the BVRI met with significant interest and regard.

We have moved from an alpha (proof of concept) version of our project to a beta one where we are in the process of testing something closer to the finished product. The current iteration of the BVRI can be found at https://omeka.library.unt.edu/s/bvri1/page/welcome. Hopefully, this work will go a long way in altering how binder's volumes have been viewed and treated in the past, and, by making them more easily discoverable, guide more research in this direction. This, in turn, will produce a more inclusive historiography that sees how music has been used in the past and how musicking took place in the nineteenth century, and one that subsequently recognizes the many types of people who participated in musical culture in a given time and place.