What’s in a genre? Metadata, Controlled Vocabularies, and the Folger’s Digital Anthology

 

Shakespeare’s plays are organized in the First Folio into three now familiar genre categories: Comedies, Tragedies, and Histories. Later scholars added a fourth, describing certain late plays like The Tempest and The Winter’s Tale that contain elements of both comedy and tragedy, along with fantastical features like magic, as “romance plays.” In organizing the 403 plays that make up the Folger’s Digital Anthology of Early Modern English Drama, we needed a few more than those four categories. By grouping our plays by genre, we aim to classify our dramatic material in ways that make it easy for scholars and readers to compare plays that were generally alike—comparing tragedy to tragedy, so to speak, using as few broadly applicable genres as possible.

To investigate the current practices for assigning genre to non-Shakespearean dramatic works, I began by consulting The Annals of English Drama, 975–1700, 3rd edition, ed. Alfred Harbage, Samuel Schoenbaum, and Sylvia Stoler Wagonheim (London: Routledge, 1989). Other major digital humanities projects in this area base their genre terms on Harbage’s Annals, including the Database of Early English Plays (DEEP) and the Visualising English Print project. I consulted Beth Ralston, VEP’s research assistant, who noted they employed a combination of Harbage’s genres (acquired from DEEP) and a much broader classification limited to four terms, “Tragedy, Tragicomedy, Comedy, and History,” assigned by research team members.

Harbage’s Annals goes with the maximalist approach, employing as many as 93 genre terms ranging from the obscure and rarely used (“mythological moral,” “Latin Satirical Comedy,” “piscatory”), to the common and familiar (“Comedy,” which according to DEEP, accounts for 456 works). 93 different classifications was an unwieldy number for our purposes. While we could eliminate some terms simply because they didn’t apply to our corpus of drama performed on London’s professional stages (for example, no “tilts”), we needed a stricter controlled vocabulary to create useful categorizations. We turned to Alan Farmer and Zach Lesser of DEEP for a model. In addition to displaying Harbage’s original terms, DEEP digested his copia into twenty-five discrete terms, so something like the “Latin Satirical Comedy” might be discoverable through searches on more general terms like “Satire” and “Comedy.” We also consulted Martin Wiggins’s British Drama 1533–1642: A Catalogue (Oxford: OUP, 2012— ), which employs terms similar to Harbage’s Annals, but pared down (and sometimes in disagreement): a total of 20 terms, ranging from “allegories” and “jigs” (like “tilts,” not part of our corpus) to “satires” and “tragedies.”

For our own system, we needed a set of genre terms that were broad enough to be flexible, yet clear enough in their definition and application to be acceptable to the Association of College and Research Libraries Rare Books and Manuscripts Section (RBMS), a division of the American Library Association. The RBMS Controlled Vocabularies provide approved, standardized terms, along with their definitions, to be used in Rare Book and Special Collections cataloging records, allowing institutions to share and compare information across collections more easily. In selecting genre terms for the Digital Anthology, we wanted to be sure we used terms that were approved by the RBMS so that our genre designation can eventually be used in Hamnet’s cataloging records for the playbooks in the Folger’s collection.

The RBMS controlled vocabulary has a section of terms specifically for plays, and when we began considering genre terms in the summer of 2015, a quick glance at the then-available genre terms for drama showed that we would need to petition for several new ones to complete the full set. In June of 2015, RBMS had terms for morality plays, masques, comedies, tragedies, and tragicomedies. Missing from that list were pastoral plays, romance plays, and history plays. Controlled vocabulary terms are added to the list based on the needs of the cataloging community: you petition the committee, a group of cataloguers considers your term and its proposed scope note (its definition) and compares the term to both already approved terms and standard dictionaries. They then tweak the definition so that it’s both clear and useful.

In August 2015 I began talking over our genre needs with Deborah Leslie, the Folger’s senior cataloguer, and Erin Blake, head of Collection Information Services. Deborah walked me through the process of submitting new genre terms to the RBMS, and by September 18th, we’d begun discussion on three new proposed terms:

• Pastoral plays, defined as “plays portraying rural life, especially in an idealized or romantic form, often with shepherds as main characters.”
• History plays, defined as “Use for plays based on historic events or well-known historical figures often from the medieval or early modern past.”
• Romance plays, which I initially proposed as “Use for plays invoking chivalric themes, including quests for love or honor, exotic locations, and heroic events. These plays often include elements of fantasy, and are relatively free of restrictive realism. Do not use for modern romance novels.”

Romance plays quickly proved to be the most controversial of these new terms. RBMS already had a term for “romances.” The scope-note for “romances” reads “Use for medieval and Renaissance tales in prose or verse celebrating the adventures, in love and war, of some hero of chivalry. Do not use for modern romance novels.”

The initial response from RBMS’s committee asked for clarification on the time span of the “romance play,” wondering if it might be the same as “romances”—a medieval as well as early modern genre. If so, one potential option would have been to amend the definition of “romances” to include drama, rather than create a new term. To clarify this, I went to Gail Gibson, a scholar of medieval drama then working at the Folger. I asked Gail if she knew of any medieval plays which fit our definition of “romance.” Gail promptly emailed back: “no.” She then elaborated on the development of the genre, the medieval history of “romances,” and the lack of pre-Reformation dramatic texts that could fit our definition, although she allowed that “the extravagant, very late (early 16th-century) Digby saint play of Mary Magdalene might be said to have some Romance-like exotic and journeying episodes.” It’s very rare that you receive such a concrete and clear answer to any kind of genre question!

The RBMS controlled vocabulary committee debated for a bit, and eventually amended the proposed definition to reduce redundant language (fantasy was seen as inherently being “relatively free of restrictive realism,” for example) and inserted a time span for which “romance plays” were written. The final definition, which is now approved, reads: “Use for Renaissance and early modern plays incorporating elements of tragedy and comedy, as well as spectacle, magic and exotic locales, and typically embracing themes of separation and reunion, forgiveness and redemption, and extraordinary adventures set in motion by the complications of love.” Sounds a lot like The Tempest, doesn’t it?

We now have eight RBMS-approved genre terms in hand:

Comedies
History plays
Masques
Morality plays
Pastoral plays
Romance plays
Tragedies
Tragicomedies

Our categories are a far more minimalist list than our comparative projects have established. We can get away with this, in part, by our restrictive corpus selection process, which required that all the plays in our collection have been performed on London’s stages. This eliminates some categories entirely, like closet dramas and tilts. Unfortunately, this also eliminated all works by women, as no women (at least currently known), wrote for the professional stage.¹ That said, the current set of genre terms doesn’t quite cover all our needs.

Ideally, all of our plays will eventually be classified as one or more of these genres. In a preliminary effort towards this classification, I translated Harbage’s copia into our controlled vocabulary. Some were easy: The Maid’s Tragedy is, quite clearly, a tragedy; the Famous Victories of Henry V, a history play; while The Careless Shepherdess is a pastoral play.

Others (in the style of DEEP), required two genre categories to accurately reflect their Harbage designation. Thus The Knight of the Burning Pestle, in Harbage’s terms a “Burlesque Romance,” became labeled as both part of “comedies” and “romance plays.”

Once these easy ones were sorted, I was left with thirteen obstinate plays that didn’t fit neatly into our scheme. These include six plays by Thomas Heywood (Brazen Age, Iron Age 1 and 2, Silver Age, Golden Age, and Love’s Mistress) which are classified by Harbage and Wiggins as “Classical Legends.”

Is a “Classical Legend” different enough from other genres to get its own RBMS approved genre? With only six “classical legends” in the corpus, I wasn’t sure our sample was big enough to warrant petitioning for it yet, but we might.

Perhaps more complex is the generic situation of what Harbage refers to as “pseudo-history” plays. The requirement that “history plays” reflect actual historic events from the medieval and early modern past means that six plays termed “pseudo histories” in the Annals cannot be easily mapped to “history plays” on our list. Wiggins has other designations for four of those plays: Jeronimo (tragedy), Nobody and Somebody (history and satire), Trial of Chivalry (romance), and The Weakest Goeth to the Wall (comedy). The other two, the anonymous Costly Whore and Thomas Dekker’s The Noble Spanish Soldier, are not yet part of Wiggins’ published catalogs. But Wiggins’s categories seem to allow us to do away with the pseudo-history genre entirely, and simply class the aforementioned plays elsewhere.

Finally, our corpus includes James Shirley’s Saint Patrick for Ireland, 1, which Harbage had classified as a “Neo-miracle” play. Since Shirley is certainly too late for the RBMS definition of miracle play—“Use for medieval religious plays based on legends of saints or martyrs, or on miracles performed by saints or sacred objects” (emphasis mine)—we’re left finding another genre for this one as well. Wiggins’s catalog is currently published up through works written before 1616; we’ll have to wait a bit to see what he says about Saint Patrick, first performed in 1639.

If anyone would like to provide suggestions for what genre we should use to catalog any of these plays, we’d appreciate it.