If you are an Anglophone who first started studying art song before the digital age, you may remember a number of classic collections of translated art song texts. My first such much-used volume was Philip L. Miller's The Ring of Words (first published in 1963), a collection of over 300 texts of mostly canonical art songs translated from German, French, Italian, Russian, and a few Scandinavian languages into English. Right around the time that I was using that book as an undergraduate pianist, however, Emily Ezust was building the original version of the LiederNet Archive.Footnote ¹ Ezust began the website as a personal project hosted on a web account at McGill University and launched it on 24 May 1995. As Ezust herself points out in the FAQs on the website, the LiederNet Archive is older than Wikipedia (Wikipedia's launch date was 15 January 2001). Since then, the site has grown (as of 17 September 2020) to include texts for choral works and to house a total of 160,284 texts in 125 different original languages, with 36,208 translations into numerous different languages posted so far. The vastness of this collection, which is constantly growing, is dependent on a global group of volunteers, who have contributed 28,625 of these translations to date. Although the site embraces choral works as well as songs, this review focuses primarily on its relevance for art song, while commenting more briefly on the website's coverage of choral music. In both realms, the database seems at first to be primarily a practical utility: a repository for the original forms of texts of arts songs and choral works, as well as a place to find translations for texts in which one has an interest. But it is in fact much more. With respect to song in particular, the resource has many features that support innovative ways of thinking about song texts in scholarship. It also serves as a kind of window into the world of art song today. The still strong pull of canonicity in scholarship and performance of the genre is palpable, even as the site offers nearly endless opportunities to go beyond the most well-known texts and songs. The website is lastly a lively platform for volunteerism and intellectual exchange. Opportunities for involvement range from contributing small suggestions concerning content to bigger translation or guest editing projects.

The database is currently searchable by ‘everything’, the surname of poets or composers (with a gender filter), title or first line, year, and collection. Some browsing features that could be useful or desirable when looking for new repertoire or patterns in the repertoire do not currently exist. For example, users cannot retrieve a list of all women composers, a list of poems in less internationally well-known original languages for art song, or – something that could be especially enlightening – a list of all women poets on the site (given the fact that a number of prominent women composers of art song have traditionally set many more poems by canonical male poets than by women). Ezust reports that users have generally not found these larger browsing features desirable, but that she is willing to collect such lists from the database herself for particular users, or to create more specialized search tools upon request.Footnote ² Specialized, experimental search tools that she has created so far on the site can be found under the link https://www.lieder.net/lieder/tools/, which is not listed at present on the main menu. There users can explore the following searches: ‘List settings by a given composer of texts written by a given poet’, ‘List all the texts that two given composers both set to music’, ‘List all settings set or published in a given year or range of years, constrained by composer or author’, ‘List unverified settings by composer or unverified texts by author’, ‘Find untranslated texts of a certain length and/or set by a given composer’, ‘List settings with missing instrument information’, and ‘List all multi-text settings’. The tool ‘Find untranslated texts …’ could be especially helpful for potential volunteers looking to contribute new translations to the site. The menu offered on the homepage of the website also includes features including an Introduction, ‘Go to a Random Text’, ‘What's New’, and FAQs, all of which provide interesting information about how the site is put together and can give newcomers a sense of its possibilities.

As Ezust notes in the ‘Special Features’ section, the site offers performers, students, and scholars the opportunity to look into the complex textual connections that can stand behind a single art song or connect different poems and songs together. Entries in the database make available information on differing titles applied to the same poem by different publishers or composers, textual variations of poems in certain settings, how composers omit or repeat certain stanzas in particular settings, the combination of different source texts in a single vocal work, poetic adaptations, settings published in various languages and the existence of alternate versions of the same original poem in different settings. Among the languages represented on the site are some from Africa, the Middle East, and East Asia, despite the fact that much of the repertoire mentioned on the site sets texts in European, North American, and South American languages. While there used to be a feature to see all of the texts in a particular language, Ezust notes that it became unwieldy, and she removed it. Users interested in such a list for languages that may be likely to have fewer entries total, however, should contact Ezust directly for help with such a search. Some of the languages on the list are also used on the site to indicate an original poem from which a published and set translation was made; in these cases, volunteers such as Sharon Krebs, and Harekrishna Meher have done important work in tracking down originals for German and other European-language song texts in languages including Zulu and Sanskrit.

Perhaps the best way to get a sense of how this collection is organized is to visit a representative page containing information and translations for a well-known and much-set poem. For this purpose, I take a cue from one of the main examples that Ezust also highlights under ‘Getting Started’, ‘Kennst du das Land’ by Johann Wolfgang von Goethe, a poem which I have also dealt with in previous research. When one goes to this entry, one finds all of the main features of an initial page for a given poem, before any particular translation on offer is readable. The full text of the poem is given in German, followed by information concerning the source of the text. In this case, the choice of a confirmation source is an interesting one, in that the site reproduces the version of the poem as found in the Goethe's Werke, Vollständige Ausgabe letzter Hand, Erster Band (1827), as opposed to the original printing of Wilhelm Meisters Lehrjahre in 1795 in which the poem first appeared. This level of detail is very useful to the individual coming to this information for the first time. A few bits of further information appear before the next standard fields on the pages, including one feature that is common across the texts on the site: the German text as presented is footnoted where well-known alterations to the text appear in well-known repertoire, in this case in Franz Schubert's setting.

The next two fields are standard to the archive and appear in the same form, where this information is available, for every text on the site. The first again concerns Goethe's authorship, this time referring to Wilhelm Meisters Lehrjahre, with the additional, green-highlighted ‘author's text checked twice against a primary source’. This standard coloured place marker exists for each authorship entry on the website and appears in red in the cases where the text has not yet been checked. The next field follows the same procedure and includes all of the known settings that have been reported to the website so far. Here, even for this well-known poem, there are still many opportunities for volunteers to contribute additional research, not only by adding settings but by verifying the settings already listed. A number of the 68 settings of the original German listed so far still contain the marking (in red) ‘sung text not yet checked’, and are waiting to be confirmed by anyone who has the actual score at hand. (Upon examining this page, I personally discovered that the Berlin-based early-nineteenth-century composer Carl Blum's operatic setting of the text has not yet been checked against the actual score. Since I discuss the setting in my own research, my next task, after writing this review, will be to contact the site with that new information.) Perhaps the most interesting information to be found about existing settings of this poem is a special section – not so typical for other entries – involving settings of this poem in other languages that have so far been reported to the site. These include settings of translations and broader adaptations by other literary luminaries, including a song on the Polish adaptation by Adam Mickiewicz, set by Stanisław Moniuszko. The links to these other translations and adaptations could provide a fascinating resource for individuals looking to trace the cross-cultural travels of Goethe's poem, even into corners not thoroughly discussed so far in the vast literature on the text.

To find the translations, the user simply has to click on the associated language codes at the top of the page; in this case, Ezust and volunteers have so far only contributed translations in Catalan, Dutch, English, French, Greek, Hungarian and Italian. As is the case with a number of poems on the site, ‘Kennst du das Land’ actually includes two different translations in some of the languages, in this case in English and Italian. In both cases one aims to be more literal, while the other seeks to reproduce some of the metre and rhyme of the original. Ezust notes that she also welcomes the addition of singing translations to the site wherever these are available or can be newly written in good quality, but these are more of a rarity. Given the large amount of volunteerism involved in these translations, the quality of translations is largely left up to individual volunteers, except for cases where Ezust or other major contributors have strong competency in the original and target languages involved and can do some editorial checking. The two English translations of ‘Kennst du das Land’ are both apt. They perhaps also show some of the struggles involved in coming up with a sufficiently original translation that doesn't follow too closely a number of earlier and more recent ones that are still under copyright elsewhere. The name of the translator is given in each case, as well as information as to how to obtain reproduction rights for the use of the translations in concert programmes and publications. (Free permission is usually granted when the purpose is non-commercial.)

With respect to this very well-known and often-set poem, it is worth comparing the information now available on the LiederNet page with that of other related databases, such as that found on the webpage for Oxford Lieder, a regular lieder festival that includes young artist training and public scholarship events in Oxford, UK.Footnote ³ In this case, the database is just a small feature within the larger web presence of Oxford Lieder. After visiting the main page of the database, the user should choose ‘explore’ among the menu options to find information about song texts. There is no clear search option for first lines, so choosing to search under ‘song’ for ‘Kennst du das Land’, while it will certainly find entries, could potentially exclude other titles like ‘Sehnsucht nach Italien’ (the title of the Blum setting mentioned above) which have appeared in the repertoire. The only four songs that appear in this search include one Schubert setting, two Robert Schumann settings, and one Hugo Wolf setting. While this very narrow list has its uses, in that these are very often performed, it is much less useful for researchers or for singers looking to expand research or programming beyond the narrowest canonical choices. There is only one (previously published, high quality translation) offered into English; no other languages are represented. The most interesting feature of this entry is that each song links to performances done at the festival itself, the factor which in fact leads to the limited entries for the text. For someone studying programming at this and other festivals, this feature could indeed be useful. However, the limited focus on just songs performed in Oxford, as well as some of the other limitations mentioned, makes this database much less useful for a number of other research and performance-planning purposes.

LiederNet, on the other hand, can also be a place for thinking through the international fame or status of songs, cycles, and the poems that they are based upon. For example, a look at the translations on offer for Die schöne Müllerin, either in full or in part, shows just how wide an impact the work continues to have. Even if that impact is limited by the status of art song as a relative niche genre in comparison to popular music or opera, the range of translations that have been contributed suggests a wide international community. The languages available for the whole cycle include English, Dutch, French, Italian, Catalan, Korean, Greek, Finnish, and Portuguese, and languages for individual songs include also Irish, Hebrew, and Norwegian. To the extent that major female song composers are also featured (Fanny Hensel is an example), the presence of translations for their song texts also tells us about the important role that male canonical German-language poets had within nineteenth-century women's song oeuvres, as well as the outsized importance that songs based on those texts – even when they can be attributed to a woman – still have today. For example, the majority of poems by Goethe or Heinrich Heine that Hensel, and often many other composers, set, link to translations in at least a few languages from the list of songs found under her name in a search. But if one seeks to find a translation of texts by Hensel's female friend Friederike Robert (1795–1832), which serve as the basis for many of her songs, or by Hensel's husband Wilhelm Hensel, whose texts she also set, one searches in vain.

This sort of a lacuna can be looked at more positively as just an opportunity for more volunteer work to continue to build the database. One rich entry into the lists of texts in their original language pertains to the songs of the African-American composer Florence Price (1888–1953), whose works have been gaining newfound attention since the discovery of a number of her unpublished manuscripts in 2009. At this point the LiederNet Archive contains a list of 16 song texts, none of which are yet translated, set by Price, including a few by the Harlem Renaissance poet Langston Hughes which cannot yet be posted due to copyright. But the texts that are there present an opportunity for translators who wish to contribute to sharing Price's songs with the non-Anglophone world, all the more so in that they involve African-American poetry that uses aspects of African-American English as spoken in the late nineteenth and earlier twentieth centuries, including by Paul Lawrence Dunbar and by Price herself. Given that such texts could be difficult for non-native speakers of English to fully understand in concert, translations into the native languages of listeners could be a helpful means of championing this repertoire outside of the US.

While LiederNet does note that it is also a repository for choral texts, a quick browse for a few representative works from different periods suggests that it is more often used by song-focused performers and scholars than by those focused on choral music. One key example of this issue concerns the Giovanni Battista Guarini text ‘Tirsi morir volea’, a madrigal text set very often in the later sixteenth century and shortly thereafter. While this madrigal text may not be one of the absolutely most performed today, it has enough fame to be regularly brought up in courses discussing sexuality in madrigal texts (the two lovers are ‘dying’ in a very coordinated way) and certainly still receives new performances and recordings with some regularity. For all of its notoriety, the text is barely addressed in LiederNet. The only translation is a German one, and no one has yet verified the text in the scores of the rather short list of compositions below. This lack of attention stands in strong contrast with the status of the poem in the ChoralWiki, a site that goes beyond text and settings information to offer scores that are in the public domain.Footnote ⁴ For ‘Tirsi morir volea’ the site gives translations of the poem in English and Dutch, and lists 13 different settings, including one recent one. It also offers numerous PDFs of the main late sixteenth-century repertoire. On the other hand, a search for something less-often performed from the nineteenth century – Robert Schumann's Szenen aus Goethes Faust – brings up no full entry at all on the ChoralWiki, and a rather full, if still underdeveloped, entry on LiederNet. These differences are perhaps fodder for research concerning the different repertoire and canonical priorities of communities surrounding choral music as opposed to lieder.

The vast collection of entries now held in the database for LiederNet should make it appealing to those scholars pursuing digital humanities methods using larger data sets. As Ezust indicates, however, the site is not currently set up so that external users can easily do this kind of work without her collaboration. If one searches ‘Everything’, for example, the search tool used is Google (albeit directed at the specific site), and many irrelevant results might appear. At this point in time Ezust suggests contacting her for developing queries to be specifically run on the database itself (and not via the public-facing interface), which can then be transmitted as a .csv file, for example. Additionally, if researchers or performers wish to search songs by subject, or the ethnic identity of composers, they might try working with the more recently established Song Helix database administered by the University of Utah.Footnote ⁵ For certain queries this site is promising; a search for all songs with the box checked for African-Americans so far turns up 198 entries.

As Ezust notes on the site, she has been its sole full-time administrator and programmer since its start, with the exception of help with a design change for the appearance of the site created by Shawn Thuris in 2019. (This site is perhaps still not a thing of great visual beauty, but that hardly matters given its success in continually accommodating changing content.) For those interested in the coding and database specifics, the coding language for the website is Perl 5, and DBIx::Class is used to connect to the Aurora (MySQL) database. The site is at present run independently of any supporting institution, and Ezust works without a salary. Revenue to support the operating costs for the site comes from donations, advertisements on the site, and licensing fees for reprinting translations in commercial situations. In the future, additional subscription-based features might be added, but the central plan is to keep the main features of the site free for users. Ezust is, however, open to clearly defined sponsorship opportunities, including from universities; readers in a position to lobby for such institutional support, especially for the purpose of ensuring the long-term stability and growth of the site, should be in touch. Ezust is to be commended for her innovative vision and her long-term leadership of the project, which deserves more generous financial support so that she can be salaried and so that the project will outlast her tenure. Given its vast scope and great potential for continuing refinement and growth in both the quantity and precision of the information offered, the LiederNet Archive has become and will hopefully continue to be an invaluable resource for anyone involved with art song and choral music. It is my hope that this review will spur increased volunteerism for the archive from scholars – especially those who can add language skills not yet fully represented on the site – and the possibility of increased long-term financial support for the project as well.