the information gleaned from spectators’ visual experiences in nonoperatic performances of song. In my hypothetical example of the angry-looking singer performing “Ich grolle nicht,” the story fact generated by the singer’s facial expression is better glossed as “the character is angry” as opposed to “the character is grimacing.” Similarly, if the singer has a beard, that does not make it appropriate to imagine that the character does. Notice the disparity with opera performance, in which a bearded, grimacing singer playing the role of Otello makes it true in the story that Otello has a beard and is currently grimacing. In an opera performance, what singers look like and the actions they perform typically generate story content about the characters they play. Likewise, the visual appearance of the stage typically generates facts about their environment.
At this point, one might raise the objection that my account describes only naturalistic approaches to stage direction. Even “traditional” productions can pose difficulties due to color-blind casting. In the Metropolitan Opera’s 1989 video recording of Otto Schenk’s production of Die Walküre, the African American soprano Jessye Norman and the white Heldentenor Gary Lakes appear as the Wälsung twins, Siegmund and Sieglinde.17 Clearly, when Sieglinde comments that she sees in Siegmund’s face a likeness of her own (Act I, Scene 3), we are not to imagine that she is lying or deceived.
The casting of Lin-Manuel Miranda’s musical Hamilton (2015) also complicates the present account, though for different reasons. It too features nonwhite singers playing white characters, but while the Met’s decision to cast Norman was made irrespective of her race, singers’ racial identities were taken into consideration by the casting directors of Hamilton. As a result, Hamilton’s casting policy is more accurately described as color conscious.18 Singers were chosen specifically because their race differed from that of their characters. These disjunctures were integral to the work’s point. In the words of the director Thomas Kail, Hamilton is “a story about America then, told by America now.”19 By casting predominantly African American and Latinx actors as America’s founding fathers, Miranda and Kail draw attention to the diversity of contemporary America and, by contrast, the lack of diversity of most of what one sees on Broadway. Daveed Diggs, who created the roles of the Marquis de Lafayette and Thomas Jefferson, describes “walk[ing] out of the show with a sense of ownership over American history. Part of it is seeing brown bodies play these people.”20 Casting Diggs as Jefferson, or Norman as Sieglinde, did not make their characters black. Yet if one were to ignore Diggs’s race, as one is encouraged to do in Norman’s case, one would miss one of the show’s key artistic and political points.
To be clear, I have no intent to discourage color-blind or color-conscious casting or to endorse discrimination against singers whose bodies do not correspond to the conventional body image of their characters.21 Rather, my point is that spectators don’t automatically assume that everything they see on stage represents contents or occurrences in the fictional world. A concern for equal opportunity for all singers, regardless of race or body type, should trump any mild discomforts that might arise from casting singers of color in “white” roles. After all, we tolerate design and direction choices that create even more blatant conflicts with the libretto and score readily enough, even when they lack the kind of ethical motivations behind color-blind and color-conscious casting.
A case in point is Martin Kušej’s Don Giovanni for the Salzburg Festival (2002). By presenting audiences with an entirely white set, frequently populated by a dozen or so women underwear models, Kušej is not inviting us to imagine that Don Giovanni’s world is literally devoid of color or that the women in this world stand about like living mannequins in nothing but their underclothes. Even though these features of the set and costuming do not represent what Don Giovanni’s world looks like, they still play a role in our understanding of the opera’s story. One way of interpreting these features is as representations of Don Giovanni’s experience of the world. The bare, colorless set could be understood as conveying his boredom and loneliness.22 The underwear models may indicate that Don Giovanni perceives women as sex objects or that this is an attitude generally held in his society.
There are also cases where the visual elements of the performance generate no story facts but merely express the director’s attitude about the work or other topics. Another way of interpreting the underwear models in Kušej’s production is to regard them as representing the director’s belief that Don Giovanni represents women as sex objects. Even under such an interpretation, the visual features of the performance still help determine its content. And the expectation that what one sees will indicate something about the visual appearance of the characters and their fictional world remains appropriate, even when the expectation is denied.
Operas versus Plays and Films
So far, everything I have said about opera also holds true of plays and films containing singing, such as Shakespeare’s Othello (1603) and the film Casablanca (1942). Despite Desdemona’s “Willow Song” and Sam’s rendition of “As Time Goes By,” neither Shakespeare’s Othello nor Casablanca is an opera. The difference between Shakespeare’s Othello and Verdi and Boito’s Otello (1887) is not merely a quantitative difference in the amount of music. It also amounts to a difference in kind, specifically one concerning the role music plays in presenting the story. Saying that an opera’s story is told through its songs and other musical numbers is a truism. In this final section, I will attempt to be more specific about what this truism might mean.
It may be tempting to claim that what differentiates songs in operas and musicals from those in nonmusical plays and films is that the former advance the plot whereas the latter are incidental to it. However, a song may be integral the plot of a play or a film (as “As Time Goes By” is to Casablanca), and many songs in operas merely provide additional insights into what a character is feeling at a given moment, without a noticeable advancement toward or away from that character’s goals.23
Perhaps songs in operas generate new story facts, whether or not these facts advance the plot. The litmus test would be whether omitting the song would result in a noticeable gap in the story. However, many songs don’t even deepen our understanding of the characters. Many merely involve characters expounding on the current situation or feelings we already know them to have, and to opera enthusiasts, such songs are no worse off for their putative superfluity.24 Furthermore, the requirement that songs generate new story facts is hardly exclusive to opera. When Ophelia sings her mad songs in Shakespeare’s Hamlet, her performances make several things true in the story, including the crucial fact that she is mad.
I suggest that one difference between operas and nonmusical plays is that song is one of the main ways characters communicate in operatic fictional worlds.25 As such, operatic worlds are strikingly nonnaturalistic, at least in this respect. When watching a play, one expects there to be a reason for a character to break into song—the character is a professional singer, for instance, or she is insane. In an opera, such explanations are not required. Songs happen anywhere at any time, even in the most unlikely scenarios, such as when one is dying of consumption!
Readers unaware of current discourses on the nature of operatic communication may find this proposal uncontroversial, even banal. However, within opera studies, the proposal that the characters are singing and, generally, hearing the music they and others make is highly contentious, or has been since the publication of Carolyn Abbate’s Unsung Voices (1991). Abbate floats two different positions on this issue. First, she states that opera characters “often suffer from deafness; they do not hear the music that is the ambient fluid of their music-drowned world.” Further along in the same paragraph, however, Abbate puts forth the more radical idea that the “music is not produced by or within the stage-world.” Developing this second proposition, she invites readers to entertain the following thought experiment: “Suppose that while attending a performance of Tosca you are suddenly transformed, given the musical ears of an operatic character. You are struck deaf to most of the singing; everyone merely speaks—except at certain moments, during the offstage cantata in Act II, when you are able to hear the phenomenal performance.”26 According to Abbate’s initial proposal, the characters are