Nineteenth-Century Women in Music: MuGI, Sophie Drinker, Art Song Augmented and BID

This article focuses on four digital resources for researching nineteenth-century women in music. Two are over two decades old and hail from the German-speaking world: Musik(vermittlung) und Gender(forschung) im Internet (MuGI) and the Sophie Drinker Institut online lexicon. The remaining pair are emerging digital resources based in the United States: Art Song Augmented and the Boulanger Initiative Database (BID).

Every project has a context for its creation and a set of priorities that shape how it is formed and/or how it has developed or shifted over time. I endeavour here to keep those contexts at the forefront of the discussion, making clear for whom or what these resources are designed and discussing their relative strengths and weaknesses in that context. While each resource has pros and cons in terms of content, layout, user experience, potential longevity, robustness and accessibility, all four provide valuable data and a wealth of information. They are all of particular interest to scholars, pedagogues and performers interested in diversifying their purview and repertories to include historically marginalized figures in music, particularly women.¹

MuGI

The Musik(vermittlung) und Gender(forschung) im Internet (MuGI) initiative was founded over two decades ago, in 2001, by groundbreaking gender scholar Beatrix Borchard, and was first based in Berlin at what is today the Universität der Künste.² The first step in what was conceived from its earliest hours as a multimedia, multimodal, digital initiative was the production of a CD-Rom about Pauline Viardot-Garcia, the famous French singer and composer. The following year Borchard was named professor at the Hochschule für Musik und Theater in Hamburg (HfMT)³ and MuGI has been centred there since, although it is currently expanding its base. Nina Noeske, head editor of MuGI along with Beatrix Borchard since 2015, recently took a professorship at the Hochschule für Musik Franz Liszt in Weimar and is continuing her work with MuGI from there, with Silke Wenzel enhancing the Hamburg-based leading team as another main editor since 2022.⁴ In addition, there is burgeoning cooperation with the Luxembourg-based CID Fraen an Gender organization, which likewise began in 2022, and is designed to unite gender research from across Europe.⁵ This broad institutional support bodes well for the future of MuGI, which, like all digital resources, will require continuous support in curating, migrating and preserving its data and access to it over time.

Conceived as a reaction to a classical music historiography focused primarily on male composers and their works, MuGI set out to reframe the history of music as a tightly interwoven cultural activity. This approach highlights the intricate networks in which women were essential, as performers, educators and promoters as well as composers. An essential part of this attempt at reworking how music history is both conceived and studied is the digital lexicon, MuGI's initial focus and still the most valuable resource it currently provides for scholars, particularly those not based in Germany. The first articles went online in 2004,⁶ and today the lexicon includes around 600 articles on (mostly) women in music, 350 of which focus on musicians active in the nineteenth century. Many featured figures are composers, but eighteenth through twentieth century women performers, publishers, scholars, sponsors, promoters, educators and patrons are all included.⁷ These include, for example, librettist, actress, estate manager, client, editor and founder of both a publishing house and a charitable trust, Gertrud Schönberg (1898–1967) who emigrated to California; German musicologist Amalie Arnheim (1863–1917); and Austrian violist, violinist, author and teacher Natalie Bauer-Lechner (1858–1921).⁸ In addition, numerous entries on men in music are also featured, including Guido Adler and Johannes Brahms, each handled explicitly from a gender studies perspective. In particular, the circles and students around Franz Liszt, Joseph Joachim and Clara Schumann were systematically researched, and a great deal of attention is also given to English musicians in the nineteenth century, so this resource is of essential value to scholars focused on those areas.

At first Lexicon articles were primarily authored by the small team of musicologists employed by MuGI thanks to funding by the Mariann Steegmann Foundation, but they were soon largely sourced out with staff fulfilling primarily editorial and publishing functions. Some 150 different scholars, both emerging and established, have contributed entries to date. Each submission undergoes thorough quality control, including extensive critical review by two internal musicologists who often go through several rounds of suggestions and corrections with the author prior to publication. Articles must conform to a rubric developed by the team which includes information about where and in which of a number of sets of capacities each subject worked, their works and output, critical reception, honours and prizes, biography and source lists. The rubric also includes a reflection on the areas in the article that require or prompt further research. They also include editorial information, such as when the articles were written and updated, recommendations on proper citation, links to authority data and other metadata. Each subsection of the rubric can also be accessed directly by clicking on the menu on the left, expediting navigation in longer articles, as shown in Figure 1. New entries are still consistently added, with interested potential contributors welcomed to get in touch with the editors.⁹

Fig. 1 Article body ‘Ernestine de Bauduin’ with floating menu on the left

The organizational rubric also partially corresponds to how the data can be filtered during searches. Although the quickest way to find a specific name is by clicking on the button ‘Lexikon’ either on the menu bar at the top or in the middle of the landing page, where names are ordered alphabetically with shortcuts available, organized by the first letter of the family name, there are also keyword [Schlagwortsuche], name [Namenssuche] and full text [Volltextsuche] options under the search button [Suche] in the top menu, all of which work efficiently. Key words are focused on geographical location (i.e. Bratislava), instrument (i.e. Violine), music type (i.e. Schauspielmusik) or cultural activity (i.e. Museumsgründerin). These varied options are particularly critical to gender research, of course, as women in music often have several different names under which they might be traced, including maiden, married and pen names.

This does bring us to the obvious challenge for many scholars using MuGI for research, which is that it is German language centric. Although some 10 per cent of MuGI articles are now available in English translations and the team is dedicated to increasing that number, the search functionality is exclusively German, and the multi-language plugin has not been set to serve the German language content when there is not an English article available. This means a few things, the first being that if the browser is set automatically to English, or if the user is searching under the English language setting [EN] instead of [DE] menu option, most searches for existing figures or articles prove less than satisfying.

More critically, the keyword search, even in English, will simply not provide results without German language search terms, likely because not all multi-language site plugins support translations of keywords or categories. Searching for ‘piano’ in the lexicon's keyword search yields nothing, while the German equivalent, ‘Klavier’, produces 266 results, neatly and alphabetically organized and with their corresponding articles just a quick click away. The full text search option is much more accommodating, and also provides brief excerpts of how the term is used in contexts. A scholar researching Robert Schumann's network, for example, might search under ‘Schumann’, and can quickly see exactly which Schumann is cited in each article, thereby avoiding wasting time on articles that will prove irrelevant (see Fig. 2).

Fig. 2 Excerpts from the first page of search results from a full text ‘Schumann’ search

The strength of the lexicon is the quality of the research within it, which is excellent, and the strengths of the site itself are its bulk, navigability (for those comfortable with German or willing to make the effort to translate), longevity and robustness – not only is the site still running after 20 years, but it is also being improved upon and expanded. After the death of Adobe Flash was announced in 2010, MuGI was given a full facelift. The system, hosted at the University of Hamburg, now runs on the MyCoRe (My Content Repository) open source framework. MyCoRe was developed and based in Germany and is specialized in discipline-specific or institutional repositories and archives.¹⁰ It is designed along current best practice for longer-term digital preservation, which includes integrating metadata models, the use of Checksum to ensure data integrity, and supporting common interoperability standards. In addition, the site is fairly responsive, meaning its appearance adjusts to the device (PC, mobile or tablet) of the user.

Another aspect of MuGI that could add to its considerable value over time is its extensive material collection. Although its 23 full multimedia expositions, originally programmed in Flash, have been archived elsewhere since the digital overhaul, they are seen as the second key leg of MuGI and are planned for reintegration into the site, under the ‘Multimedia’ banner on the landing page. Data migration and curation, costly, necessary and time-intensive processes, have recently been given a shot in the arm by Gen-AI, which can now be leveraged to powerfully expedite these processes, making batch processing more accurate as well as quicker and less costly.¹¹ In the meantime, many of the lexicon articles already integrate licensed scores, sound samples or full audio recordings as well as visuals. The team is currently looking to expand and systematize their material collection in the interest of providing a link between the existing biographical lexicon and a planned, second set of gender-studies articles, the Sachteil, focused on non-biographical subjects of interest.

Sophie Drinker Institut

Although similar in certain ways to MuGI and with a degree of overlap, Sophie Drinker has a narrower purview and is the most ‘complete’ digital collection of the four laid out here.¹² Focused on European, female instrumentalists of the eighteenth and nineteenth centuries, the lexicon of instrumentalists [Instrumentalistinnen-Lexikon] of the Bremen-based Sophie Drinker Institut is the ne plus ultra in online scholarship. Boasting circa 750 biographical articles of musicians born in Europe prior to 1880, it is the sole resource of its kind to prioritize female performers – instrumentalists specifically – rather than composers or singers, reasoning that even more eighteenth and nineteenth century women were active performers than composers, and are even more absent from music history. ‘Female instrumentalists are therefore of historical interest as pioneers of female professional activity’, the initiative explains, continuing that ‘one of the aims of the instrumentalists’ lexicon is to bring these often forgotten pioneers back into historical memory’.¹³

The institute itself, like MuGI, was founded in 2001 as an independent research institute dedicated to gender and women's studies. It began its online lexicon output in 2006, but has shifted its focus over time to include cultural studies aspects of how music histories are created and retold, laying out how educational structures and professional networks, for example, functioned historically. The centre publishes large-scale anthologies¹⁴ as well as a regular series of similarly gender-studies-focused monographs, put out primarily by the institute's founder, musicologist and gender studies scholar Freia Hoffmann.¹⁵

A slight difference to MuGI is the percentage of articles created by about a dozen of the institute staff which is higher at Sophie Drinker, and the articles are, generally speaking, shorter. Creating the instrumentalists’ lexicon was the institute's singular focus for around a decade, and the bulk of the articles were written by 13 core staff members,¹⁶ although both external specialized scholars and interested students were also asked or allowed to contribute. Those articles were then edited by staff scholars, following a specific classification and style structure designed for the lexicon.¹⁷ In around 2015 the focus shifted, and although new articles still trickle in today, the bulk of the energy is now directed towards related yet distinct research endeavours. They published a three-volume handbook focused on music conservatories in 2021, for example, and many of the more recent volumes published in their own series are likewise focused on eighteenth and nineteenth century music pedagogy and networks.¹⁸

The institute also has an extensive physical and digital collection which includes recordings, materials from personal estates and partial estates, literature and music scores. Online access is only available to students at the University of Oldenburg, however, while the physical collections housed in Bremen have not yet been digitized. That being said, external scholars with questions or those who are looking into resources that may be located at the institute may contact the team members, who are generally helpful and more than accommodating. Through such inquiries scholars can certainly find out in advance if what they are looking for is worth a physical trip to Germany.¹⁹

From a technical standpoint, although the website was completely renovated a few years back, it is still slightly daunting to non-German speakers. The quickest option is simply to use the general text search option [Suche] in the upper right-hand corner of the landing page (see Fig. 3). Alternatively, clicking on ‘Instrumentalistinnen-Lexikon’ (see Fig. 4), the furthest menu option to the right, yields not only an extensive introduction to the lexicon's concept and contents (the same content as in the submenu Lexikon) but also the option to browse articles alphabetically, by family name. In addition, a valuable glossary of German abbreviations is available in the submenu Abkürzungsverzeichnis and guidelines for authors in the submenu Richtlinien für AutorInnen. Beware, these submenus are no longer accessible from the page Europäische Instrumentalistinnen des 18. und 19. Jahrhunderts; they disappear from the top after the user has entered the lexicon itself and become available in the menu to the left.

Fig. 3 Sophie Drinker Institut landing page

Fig. 4 Dropdown menu for the lexicon

Unfortunately, there are no search filters and no way to narrow results by categories or classifications like location, instrument or other keyword types; someone looking to quickly find a number of Romanian keyboardists, for example, will have to do it the hard way. In addition, within articles there is no syllable-break function which can be slightly annoying to readers sensitive to visualization, and unlike MuGI, no subsections or chapters within each article to which the user could jump directly. That being said, the vast majority of the articles are concise enough not to warrant subdivisions.

As with MuGI, the Sophie Drinker lexicon is in German, but unlike MuGI, their articles will likely remain exclusively so. Although enough reviews and literature included are in English that one might get lucky through the full-text search, and excerpts are, helpfully, provided within this framework, this is a lexicon aimed at German-speakers interested either in browsing in a very general way, or for searching for information of a specific instrumentalist or network, based on a known family name (see Fig. 5). Full text querying is quick and painless for the user who knows where s/he is going, and although this is one of the most-visited sites for gender scholars in Europe, its content delivery is more than capable of handling the traffic without noticeable lags. In addition, like MuGI, the quality of the research and information found in these articles is exceptionally high and it is regrettable that it is not yet a standard source in the English-speaking world.

Fig. 5 Full text search for ‘Lang’

In terms of further digital assets provided by the institute via the site, there are valuable links to downloadable PDFs. For example, under the menu option Zum Nachschlagen under ‘Materialien zur Geschichte der Konservatorien’ there are multiple links leading to a wealth of information about nineteenth century music conservatories in the German speaking world.²⁰ Collections of secondary literature, lists of students, graduates and when they began studies, and information about what these educational paths entailed are sorted by city, providing a rich view into educational systems that fundamentally shaped classical music education. There is likewise a multipage document under ‘Orchesterwerke von Komponistinnen’, rich in content, which includes a list of all available orchestral music – 588 works total at the time of writing²¹ – by women composers that interested orchestras could easily rent or buy, compiled in 2012 by ForumMusikDiversität;²² and under Publikationen there are four articles published by the institute which are made freely available online, covering topics ranging from the historical definition of a vocal soubrette to the extensive career paths for female professional orchestral musicians.²³

Art Song Augmented

Compared to the German language monoliths covered above, Art Song Augmented, founded as a passion project by Stephen Rodgers, a professor of music theory at the University of Oregon School of Music and Dance, in 2021 is a tiny newcomer to the scene.²⁴ Its attractive layout, low context design, pedagogical focus and prioritization of high quality audio-visual recordings make it particularly valuable for young scholars, as well as students, performers, and vocal or collaborative piano teachers looking for inspiration to diversify their repertory arsenal.

The impetus for creating Art Song Augmented was a realization that both the quality and number of recordings of works by marginalized composers left significant room for improvement. Assuming that students fell in love with new repertoire primarily through hearing – and seeing – it performed at a high level, Rodgers reasoned that publication and information about unknown or little-known works would be insufficient to shift performance repertories or syllabi organically. He launched the initiative, equal parts embedded video performances published via YouTube and miniature biographical sketches or analysis, enhanced by links to scores and further literature, when available. The initial pool of five composers, which are tagged geographically, has grown to 26 at the time of writing, 18 of which are women born in the nineteenth century. The pedagogical aim of the site is evident through the little tidbits of information planted throughout the site, which encourage students to explore at their own pace, without being overwhelmed (see Fig. 6).

Fig. 6 Pedagogical snacks

Although Rodgers runs and finances the project alone, supported by funds from his professorship, and the written content is solely curated by him, the content, including numerous world-premiere recordings, is generated collaboratively. He has worked with around a dozen musicians and scholars to date, and features recordings by numerous professional as well as student singers including Julia Bullock, Rebecca Nelsen, Camille Ortiz, Kenneth Overton and Klemens Sander and pianists such as Maria Thompson Corley, Jocelyn Freeman, Myra Huang, Andrea Kmecova and William Chapman Nyaho. The design, by UK-based Anastasia Witts, founder of Artist Digital Limited, who is likewise responsible for the Women's Song Forum site, is attractive and designed to encourage scrolling and browsing in patently non-intimidating fashion.²⁵

Although ideal for student/performer browsing, the site is not a database and not optimized for searches, which will need to be rectified as its content grows. Currently it lacks any search functionality, and the landing page offers access solely by browsing a handful of composer portrait thumbnails, tagged geographically (France, Germany, the United States, England). Through a further click they can access the full list of composers and thumbnails on a single long-scroll page, not organized under any pattern, alphabetical or otherwise. An alternate click option takes the user to the YouTube landing page for the recordings, which are organized in playlists by composer names, likewise not following any sort of organizational schema (see Fig. 7). Technically, the reaction time after clicking also lags discernibly, an indication that images or audio may not be compressed optimally. This is a non-issue if the user is simply browsing and scrolling, but it would quickly become frustrating for a scholar searching for specific information or trying to jump between pages quickly.

Fig. 7 Art Song Augmented video playlists, www.youtube.com/@artsongaugmented7162/playlists.

That being said, the site is well suited for its intended target audience and is unique in its simple design and ease of use.²⁶ It is a particularly appropriate entry point for young scholars, teachers or musicians interested in discovery and exploration, and the emphasis on high-quality recordings and sleek presentation is important to give marginalized works value – many scholars, after all, can attest to being underwhelmed by women's compositions, not because of any lack of inherent quality but simply because the available recording or performance was not convincing. In addition, Rodgers specifically welcomes collaboration with further potential content creators, who are encouraged to reach out to him directly via the site's contact portal or its Facebook Messenger function.

Boulanger Initiative Database (BID)

The final resource is an emerging powerhouse from the USA, the extremely ambitious Boulanger Initiative Database (BID).²⁷ If the focus of Art Song Augmented is pedagogical, BID is specifically directed for performers looking for scores, with a healthy degree of implicit side-eye directed towards any organization which purports to be willing to diversify its repertoire, were it not for the lack of available music by marginalized composers or available scores (see Fig. 8).

Fig. 8 Mission statement and database landing page

The database is an outgrowth of an equally energetic initiative spearheaded by musician-scholars Laura Colgate and Joy-Leilani Garbutt around 2018 who began with a general interest in diversifying the classical landscape in a variety of ways.²⁸ Their team (and a blend of private and corporate financing) have grown remarkably and at lightning speed,²⁹ and their current range of activities include performance events and festivals, Wiki-thons devoted to enhancing or improving Wikipedia pages of marginalized composers, and numerous virtual events, to name just a few of their extensive portfolio activities.³⁰

The database itself was launched in March 2023 and focuses explicitly on BIPOC, women, and gender marginalized composers. It contained from the onset over ‘8,000 available works by 1,200 women and gender marginalised (“non-cis-male, including trans and/or non-binary”) composers available for free’,³¹ and has expressed plans to continuously expand both its collection and searchability over time. At the time of writing the composers included in the database include nearly 600 women born between 1780 and 1890.

At the moment users have the option to search by composer name, with the option to filter by country and US state,³² as well as by limiting both birth and death years. One suspects that further search options – by composer ethnicity, sexual orientation or gender identification might be of real interest to their users, but this is, of course, an extremely complex undertaking. The other main option is searching by work, with multiple and additive limits available, including composer name, duration or publication year (in ranges), instrumentation, composition type, publisher, language or ‘attribute’, which at the moment offers limiting for those who have publishers (or do not) or whether a recording is available. As a collaborative pianist, and considering that art song is one of the two genres within which women historically composed most extensively, I was, frankly, surprised that it was not easier to limit results to music for one voice, one piano, find a category called ‘song’ (although ‘song cycle’ exists) or find something along the lines of ‘voice/piano duo’ instrumentation within the plethora of available search options, but perhaps this is splitting hairs. There are certainly many options already available, and this is still a resource in its relative infancy.

An excellent addition is a practical guide for users, something generally absent within scholarly databases, many of which require considerable more technical savvy to negotiate.³³ Another is that searches for works regularly offer considerable information beyond the composer's name, including anthology titles, notes, links to scores (when available) including publisher's sites for purchasing scores, library holdings containing the score (WorldCat, etc.), and/or online access to PDFs as well as recording information (CDs to purchase) and links to online audio or audio-video recordings.³⁴ This is partly because of its contribution structure, which is well organized, and leverages the Airtable platform to outsource the generation of structured data which is readily ingestible into BID without much additional manual curation.³⁵

In terms of weaknesses, beyond what might prove minor annoyances to some users such as the persistent banners designed to attract advertisers and partners, it is extremely broad in what is promised, which is both a potential strength and an Achilles heel. At the moment there are an extensive number of composers only listed by name without works or further valuable information, and no ability to filter for entries where a work is attached or a score available (see Fig. 9), which goes slightly contrary to the stated purpose, i.e. demonstrating that scores are easily available. Many of the composers and their linked information do not reflect the most up to date scholarship, as sources they are scraping (Donne's big list, for example) ³⁶ are exclusively English language: although themselves large in scope they are not as thoroughly researched as MuGi or Sophie Drinker. Finally, since BID offers so much linked content and digital content is constantly in flux, the risk is real that in a few years, after initial enthusiasms and funds have dissipated, BID might well amount to no more than a graveyard of article stubs and broken links.

Fig. 9 Stubs for entries in the Boulanger Initiative Database

If it is continuously refined, updated and funded, however, BID would prove an unparalleled source. There are numerous future additions already in the pipeline, including adding living composers and their works, and pointing to archival locations for unpublished works by historical composers, and offering searches which allow for shorthand orchestral instrumentation work and filtering by difficulty level, as well as both descriptive/thematic tags. Even without, this is an important and ambitious step in a considerably more inclusive direction for classical repertories. I am pleased to see such a quality effort of this scope attempted, and look forward to observing how it develops.

Final Thoughts

As we move with alarming speed into times where scholarship and research is digitally driven, it is comforting that resources for researching women in music have been developed from the outset, and that they continue to be created to suit a variety of needs and priorities. Despite the desire to standardize, I firmly contend that resources, and especially those focused on marginalized composers who often are characterized by difference to norms, need to leave room for deviation. I also am hopeful that modern shifts will alleviate my overarching concerns, two of which emerged while looking into these resources as a group – the first having to do with accessibility and the second with language.

Although all sites were designed to be more or less responsive, none are truly accessible along design best practices. Those unable to use a mouse will not have access as keyboard navigability is largely missing, a major topic at the moment in digital design. More concerning yet, however, is what I understand as the siloing of big databases, generally along the lines of language. I hope that MuGI will continue to make their site dual-language so that English speaking scholars will have access to the best research, and I think, frankly, that Sophie Drinker should do the same. By the same token, I hope that BID takes advantage of emerging AI technology, capitalizing on its extensive language translation possibilities to improve the quality and depth as well as the breadth of their information and offering by scraping non English-language resources as well.

Finally, a brief plea for historical musicologists and humanities scholars to care about these and other digital resources, even if they do not use them personally. It is, of course, particularly critical that gender scholars care about the digital revolution, simply because the stakes are so high. The promise inherent within is an offer to level the playing field by creating powerful resources and thereby scholarly and performance practices which are much more diverse and inclusive than they have ever been in recorded history. If concerns of inclusivity and equality are at the forefront of creating research tools, as they are with these four examples, and if these tools are supported and used, there will be a paradigm shift in the research and performance communities that cannot be ignored. If not, the tendency to train AI on so-called standard data will simply recreate a digital reality dominated by white, male composers and their works, rolling back decades of feminist scholarship and progress. Data is not unbiased, and it has always been generated unequally.³⁷