Notation of operatic laughter

Among the murder and mayhem, yearning, lusting, and dying in the musical melodrama of opera, singers perform the occasional chuckle, yuk, or titter. Such theatrical emoting is data encoded in the musical score and is subject to quantitative analysis. The musical score is a set of instructions to a vocalist or instrumentalist about the duration, pitch, and shape of a sound to be produced. If you know the time signature, metronome marking, note value, and pitch, you can represent and reproduce a sonic event such as laughter with reasonable accuracy. Can we tap the fabled musical powers of Mozart and colleagues to capture the essence of laughter? Teamed with fellow musicians Helen Weems and Lisa Griesman, I set out to answer these and related questions (Provine Reference Provine2000). After considering the description of laughter via musical notation, we close the circle by evaluating the potency of notated and musical laughter to trigger contagious laughter in audiences.

The starting point of our analysis was the sound of laughter itself that composers and singers seek to emulate. Waveform and spectrographic analyses (Provine & Yong Reference Provine and Yong1991) defined laughter as a series of “laugh notes” (“ha,” etc.), bursts of sound that last about one-fifteenth second and repeat every one-fifth second or so (“ha-ha,” etc.). A sigh-like aspiration is present before the first laugh note, and between all others. If laugh notes are digitally removed from recordings of laughter and the gap closed, all that remains is a long, breathy sigh. In addition, laughs, particularly lengthy ones, proceed with a decrescendo, a progressive decrease in loudness. Our analysis produced a variety of notation schemes, time signatures, and metronome markings that accurately describe laughter (Provine Reference Provine2000). But the present focus is on existing opera scores, not our proposals for new ones.

We found 57 operas with laughter notated in the score, but focused on 20 operas for which we had both scores and at least two different recorded performances. We then mathematically converted scored laughter into laugh-note durations and inter-note intervals in milliseconds, or, metronome in hand, converted the actual performance of laughter into similar data. A wide variety of notation schemes was encountered, ranging from an occasional thirty-second, half, and whole note to the more usual sixteenth, eighth, and quarter note. The pitch of laugh notes was usually scored, but sometimes only a rhythmic marking was given. Of course, these notations have meaning only in the context of time signature and tempo. Leoncavallo avoids notation altogether in I Pagliacci, simply instructing the murderous clown Canio to “laugh bitterly” in the Italianate Ah! Ah! Ah! Ah! Ah!

Most composers, Mozart included, usually overestimated the duration of laugh notes (“ha”) and inter-note intervals (“ha-ha”). Thus, composers typically notate and singers perform laughter more slowly than it really is. In contrast to the stereotypy of conversational laughter, there is a greater range of cadence of notated laughter between composers, and even for the same composer within a song or opera. For example, in Mussorgsky's Boris Godunov, the tempi of four different laughter-bearing songs range from about 35% too slow to 50% too fast. Scoring laughter carries a built-in trap for composers; they are constrained by the tempos of a song already in full flight, and by the demands of the dramatic and musical moment. With a few exceptions, composers did not acknowledge the silent interval separating laugh notes (“ha-ha”), or laughter's natural decrescendo. In performance, some singers took liberty with the score, perhaps in the interest of dramatic context, or necessity of vocal production. The most flagrant variance was in adding laughter where none was scored, as typical in the aria “E Scherzo od e Follia” from Verdi's Un Ballo in Maschera. No cases were found of omitting scored laughter. When it comes to emoting, opera singers would rather giveth than taketh away.

At study's end, we learned more about opera than about laughter. In opera, too many concessions are made to the tyranny of the musical moment, and laughter is in the service of song, not vice versa. Thus, the fabled ears of Mozart and colleagues never got a full test and we are left musing about their analytic prowess.

Singing in the primal chorus

One of the strangest human rituals is our joining fellow Homo sapiens in the primal chorus of ha-ha-ha. We repeat the sound of laughter when we hear it, producing a chain reaction of neurologically primed levity (Provine Reference Provine1992). Similar contagion is characteristic of yawning (Provine Reference Provine2005). Like laughter in conversation, contagious laughter “just happens.” It is not a conscious choice. We do not speak “ha-ha” as we choose a word in speech. Laughter, like crying, is difficult to produce on command.

Beginning with theatrical performances in ancient Greece, contagion has been recognized and used to enhance, and sometimes to suppress, audience response (Provine Reference Provine2000). So-called claques of stooges were planted in the audience to prompt the audience into greater laughter, cheering, and sometimes jeering. Emperor Nero, an avid actor, ordered thousands of Roman soldiers to attend his performances and applaud loudly. Wise judges always awarded Nero first prize. The use of claques, formal and informal, continues in opera to influence audiences and newspaper reviewers, in political rallies, religious revivals, and other settings where a mass emotional reaction is desired.

Opera is occasion to another historical precedent: the use of contagious laughter in musical and theatrical performance. Michael Kelly, Irish tenor and friend of Mozart, was performing the delightful “Haste Thee, Nymph” from Handel's L'Allegro, il Penseroso ed il Moderato, a song featuring some sung laughter. After being unsatisfied with stiff performances by other singers, Kelly, according to his own report,

Handel and Kelly tapped an aspect of human nature that was exploited by the entertainment technology over a century later.

Contagion was the theme of the 1922 “Okeh Laughing Record,” a trumpet performance interrupted by highly infectious laughter that became one of the most effective and successful novelty records of all time (Provine Reference Provine2000). Louis Armstrong, Jelly Roll Morton, Sidney Bechet, Spike Jones, and others followed with their own variants. In 1950, television's Hank McCune Show added “canned laughter” to its broadcasts to compensate for the show being recorded in a studio without a live audience. “Laugh tracks” continue on television sitcoms because they actually do produce audience laughter and increased ratings of humorousness. The mechanism of contagious laughter and yawning may involve a specific sensory detector that triggers these acts. Contagion requires no mirror neurons.

This light-hearted musical interlude is offered in the belief that the scientific and artistic issues in music and emotion are broad, deep, and will most likely yield to unanticipated, interdisciplinary approaches, even one including opera. The target article prompted my musing about such topics as singing, laughing, yawning, chanting, poetry reading, talking, dancing, and other rhythmic, stereotyped, and sometimes contagious acts that may or may not be social. An alternative commentary was “Movements of Note,” a consideration that all vocalizations and sonic acts are movements that produce sounds, with contagious sounds being movements that cause movements in observers. I favor approaches that treat sound-making as an objective, observable act, like walking, breathing, or flying, and that avoid a long slog through the semantic swamp of music definition. We may think more clearly about music when we make it less special.