Kivy and the “Problem of Opera”

The Problem and the Key to Its Dissolution

The view that music and literature are inimical has long been associated with musical formalism. This view can be traced as far back as Hanslick.¹ More recently, Peter Kivy has given the classic argument for the existence of a “problem of opera.” The problem stems, Kivy believes, from a fundamental difference between the literary and musical elements of opera. In his view, literature is a narrative genre with semantic content, and works of literature unfold in a manner determined by semantic considerations. In contrast, Kivy tells us, music is contentless form and makes sense only in purely musical terms. He believes that composers face a stark choice when they set a libretto to music: they can compose music that is subservient to the text and enhances its semantic content, or they can write music that is successful in purely musical terms. They cannot do both. This is the “problem of opera.” As I shall argue, however, there is no such “problem of opera.” It is only thought to exist because formalists such as Kivy misunderstand both music and literature.

We need to be clear about the problem under consideration. Sometimes people speak of a problem of opera and mean nothing more than the fact that composers of opera face a problem of getting drama and music to work together. When Alban Berg used the phrase “the problem of opera” in a 1928 essay, this was the problem to which he referred. Winton Dean writes of the “problems of opera.”² An opera production is more than a combination of words and music. It also involves acting, scenery, and (often) dance. Avariety of problems must be overcome if these are to be effectively combined. For example, Dean notes that there is a problem of finding a pace that works both for music and for lyrics. This is a genuine practical artistic problem, not one of the problems that philosophers concoct. The “problem of opera” that Kivy describes is, in contrast, an unsolvable pseudoproblem.

The “problem of opera” begins with formalism about music. (I will use quotation marks when referring to the pseudoproblem as opposed to the—very real—practical problems.) Kivy believes that music is “pure, empty decoration,” comparable to a design on wallpaper, a pattern of colors viewed through a kaleidoscope or the decoration on a Persian carpet.³ In his view, music cannot contribute to the content of an opera and cannot give it psychological insight. Music is set apart from the other decorative arts only by the intellectual challenge involved in grasping the structure of the music. Opera is no exception to this rule. The operas of Monteverdi, Handel, Puccini, Wagner, and Berg are stylistically diverse, but, regardless of the styles composers of opera employ, Kivy believes that they face an insurmountable challenge when combining musical form with words. According to Kivy, an opera libretto, like any other literary text, unfolds as a series of statements with semantic content. From a dramatic point of view, it makes no sense, he believes, to adapt a libretto to the pace and pattern of a musical composition. From a musical point of view, the linear unfolding of a drama is unsatisfactory.

According to Kivy, as composers grapple with the “problem of opera,” two outcomes are possible. In the first outcome, composers put music at the service of the libretto. The music is used to assist a libretto in conveying meaning, that is, semantic content. This outcome is found in the works composed in the stile rappresentativo. Kivy regards these works as “music drama” rather than opera. Music drama is held to be unsatisfactory because the musical component of a work in this style will fail as pure music: it will lack musical interest. “After a while the stile rappresentativo begins to pall, whether it is by Peri, Caccini, or even Monteverdi.”⁴ This is a clue that the critical direction is on the wrong track: when someone finds himself writing that Monteverdi palls, the only thing to do is to go back and find where the argument has gone...