Common-tone Tonality in Italian Romantic Opera: An Introduction

I.

[1] As Clausewitz famously declared, war is a continuation of politics by other means.⁽¹⁾ By now it is hardly news that music analysis, in almost all forms inherited from the nineteenth and early twentieth centuries, is a continuation of German Romantic ideology by other means. Yet saying so (and it has been said for over a quarter-century now) has changed remarkably little in the day-to-day life of North American music theorists. In particular, it has not changed the pre-twentieth-century repertoires that we study and teach. German Romantic hegemony has been challenged in many ways—most loudly perhaps by Richard Taruskin⁽²⁾—but theorists have hardly begun to do so using other repertoires, contemporary to its ascendancy, that might offer alternatives to it. It seems foolish to eschew organicist methods of analysis for music conceived according to organicist premises,⁽³⁾ but it is equally foolish to apply only organicist methods to music differently conceived. Without seeking to topple German Romanticism within its own domain, we should explore other domains and try to view them in other ways. Only, since German Romanticism saw its domain as universal, to relativize it in this way—to deny its universalizing claims—is necessarily to dethrone it. We can continue to draw our Schenker graphs and tell our Carpenter narratives without expecting the world to submit to Schenkerian or Schoenbergian standards of measurement.⁽⁴⁾ Similarly, we can continue to cherish the Austro-German repertoire from Bach to Webern without granting it the aesthetic exclusivity that its apologists sought or, in many cases, assumed.⁽⁵⁾

[2] For each dominant culture, there are multiple countercultures. The counterculture that I choose to cultivate is nineteenth-century Italian opera.⁽⁶⁾ In this essay, intended as a prolegomenon to a larger study, I argue for the historical and theoretical importance of this repertoire within the larger corpus of nineteenth-century music. This importance lies partly in historical priority: some innovations that have widely been ascribed to German composers can more plausibly be traced to Italy. At a minimum, Italian composers discovered them independently and popularized them earlier than composers in other parts of Europe.

II.

[3] It has been thirty years since Robert Bailey introduced five concepts for the analysis of Wagner’s operas: associative tonality; expressive tonality; directional tonality; tonal pairing; and the double-tonic complex.⁽⁷⁾ The double-tonic complex remains controversial, but Bailey’s other ideas have been widely accepted, at least in North American scholarship. Directional tonality and tonal pairing have proven especially important to the analysis of Austro-German music from Wagner through Wolf, Mahler, and early Schoenberg, offering an alternative to more traditional approaches, chiefly Schenker-inspired, that seemed to work less well for the post-Wagnerian repertoire. The state of this art circa 1990 can be reviewed in the well-known volume The Second Practice of Nineteenth-Century Tonality, a Bailey Festschrift in all but name.⁽⁸⁾

[4] Bailey’s concepts have also been applied to music before Wagner. Harald Krebs has been especially active here, publishing several articles on directional tonality and tonal pairing in the music of Schubert and Chopin.⁽⁹⁾ William Kinderman and Kevin Korsyn have also discussed Chopin’s two-key pieces from a more or less Bailey-influenced point of view.⁽¹⁰⁾ Peter Kaminsky has explored the subject of tonal pairing and directional tonality in Schumann’s early piano cycles, albeit without direct reference to Bailey’s ideas.⁽¹¹⁾

[5] Virtually all of this attention has been focused on music by Austro-German composers, the sole exception being Chopin—a composer who, as most theorists know, was adopted as an honorary German by Heinrich Schenker, much as Berlioz was adopted (for different reasons!) by adherents of the New German School.⁽¹²⁾ William Kinderman, in his introduction to The Second Practice, acknowledges the Austro-German focus of that volume, while expressing the hope that scholars will apply its approach to “French, Russian, and other musical repertoires of the nineteenth century.”⁽¹³⁾

[6] David Kopp’s recent book, Chromatic Transformations in Nineteenth-Century Music, continues the Austro-German bias. Kopp’s book is largely about chromatic third-relations, especially modally matched transformations involving one common tone—for example, a move from a major triad to its major mediant (III) or lowered submediant (VI).⁽¹⁴⁾ There is nothing peculiarly Germanic about chromatic third-relations, but when it comes time for analytical illustration Kopp rounds up the usual suspects: Schubert, Chopin, and Wolf. He does include pieces by Liszt and Dvorak, two composers closely identified with German styles, and a short excerpt by Smetana. There is even a snippet of Chausson.

[7] Kopp’s omission of Italian music is both noteworthy and symptomatic: noteworthy, because in chromatic third-relations it was Rossini who set the pattern for the rest of Europe to follow; symptomatic, because the marginalization of nineteenth-century Italian opera has long been a defining feature of North American music theory. I can only sketch the relevant history here. Italian opera was the Other against which German Romantic musical culture defined itself, as can be seen in the writings of critics from Schumann to Wagner, Hanslick, and Schenker.⁽¹⁵⁾ Even before the arrival in America of Schoenberg, Hindemith, and the Schenker circle, those American musicians involved in teaching and writing about music theory tended to be Germanic in their musical sympathies, and some (including Percy Goetschius) studied in Germany. But other, non-Germanic viewpoints continued to flourish—as evidenced, for example, by those American composers who flocked to Nadia Boulanger. The musico-intellectual climate changed gradually after the arrival of Austro-German émigrés, who impressed their Germanocentrism on American composers and, still more profoundly, on American theorists.⁽¹⁶⁾ That theory was even more deeply affected than composition surely owed to the fact that speculative theory was overwhelmingly a German field after 1850; to think theoretically was, almost by definition, to think Germanically. By the 1950s American theory was dominated by an Austro-German (often specifically Viennese) aesthetic, which only increased as professional theorists progressively displaced composers in the theory classroom. While American opera fans were thrilling to Maria Callas and the bel canto revival, elite opinion held that the only acceptable Italian operas were Otello (Verdi’s, not Rossini’s), Falstaff, and anything by Mozart. The only nineteenth-century vocal pieces encountered in the typical theory classroom were, and are, Lieder.

[8] What, for an earlier generation, was principled Germanocentrism has continued to the present day through a combination of inertia, indoctrination, and prejudice. While twentieth-century popular music has been declared ripe for theoretical attention, American theorists have largely cleaved to the “classical” repertoires they learned in their youth and the prejudices they imbibed from their teachers. I speak, of course, partly from personal experience.

[9] As I suggested earlier, many of the innovative practices addressed by Bailey, Kopp, and others can be found in early-nineteenth-century Italian opera, which may represent their historical source. That Schubert was affected by the Viennese mania for Rossini; that Chopin’s melody shows the imprint of Rossini and Bellini; that Wagner admired Bellini’s melodies and his dramatic sense—all of these things are well known. The effects of Rossini’s music on Schubert’s harmonic practice are only beginning to be described,⁽¹⁷⁾ while Italian influence on Schumann seems to have gone unrecognized. Myths created by Wagner himself have obscured recognition of what he learned from his Italian predecessors.⁽¹⁸⁾

III.

[10] The same decades that saw the efflorescence of North American scholarship on Wagner and his successors saw an unprecedented upsurge of analytical attention to the music of Verdi. An early indication of both trends was the 1984 Verdi–Wagner conference at Cornell University (which I did not attend) and the book, Analyzing Opera, that grew out of it.⁽¹⁹⁾ The editors of that volume, Carolyn Abbate and Roger Parker, juxtaposed Verdi essays and Wagner essays as carefully as a chief of protocol seating guests at a state dinner, but they failed to hide the lack of scholarly intercourse between Verdians and Wagnerians. Like the Italian and German divisions of the Metropolitan Opera in its early decades, the two groups of scholars merely occupied the same building. They may have chatted cordially, but they didn’t cite each others’ work—this despite the shared heritage of their heroes in the music of Spontini, Rossini, Bellini, Halévy, Meyerbeer, et al. The situation was different in Europe, where a similar conference had been held in 1969 with far greater emphasis on the mixing of national traditions.⁽²⁰⁾

[11] The new Verdi scholarship has been extremely impressive, yet this literature seems to be unknown to all but a few American music theorists.⁽²¹⁾ Although much of this work has been carried out by Europeans (among them Pierluigi Petrobelli, Friedrich Lippmann, and Gilles de Van), many leading Verdi analysts have been North Americans; yet hardly any have held professional positions in music theory.⁽²²⁾ Verdi analysts to emerge in the 1970s included Martin Chusid, David Lawton, and Robert Moreen. The ‘80s and ‘90s produced Harold Powers, James Hepokoski, Scott Balthazar, Steven Huebner, and Roger Parker (although Parker has become more of an anti-analyst). Among the younger generation there is Andreas Giger, who specializes in Verdi’s French operas. Every one of these figures has emerged from the disciplines of historical musicology or, in Powers’ case, ethnomusicology. To ask why would be to confront some of the most deeply entrenched biases of American music theory.

[12] Let us turn instead to some of the positive ideas that have emerged from Verdi scholarship. A great deal of attention has been paid, understandably, to matters of musical form, since formal principles in this repertoire differ significantly from those familiar to theorists from the study of Viennese instrumental music ca. 1775–1815.⁽²³⁾ More important, for my purposes, has been an idea that was proposed, apparently independently, by Pierluigi Petrobelli and Martin Chusid in the early 1970s. Petrobelli argued that Azucena’s music in Il trovatore centers around a single melodic pitch, B4, which is alternately harmonized by the keys of E minor and G major.⁽²⁴⁾ At about the same time, Chusid demonstrated the centrality of the pitch C4 to the music sung by two baritone characters in Rigoletto, Monterone and Rigoletto himself.⁽²⁵⁾ In effect, Petrobelli and Chusid argued that, at least for these characters, the center of musical gravity lay not in the orchestral bass but in the vocal line. Instead of emanating upward from a fundamental, as in the tonal theories of Rameau and Schenker, musical coherence emanates downward from a focal melodic pitch. Petrobelli dubbed such a focal pitch a sonorità (“sonority”).

[13] The concept of sonorità was adopted by other scholars, including Harold Powers and David Rosen.⁽²⁶⁾ Both promoted the sonorità as a distinctively Italian organizing principle, arguing (as I shall argue) that Verdi’s music follows principles different from those that underlie most Austro-German music. In addition to building on Chusid’s work on “baritone C,” Powers offered an analysis of the love duet from Verdi’s Otello, also based on the sonorità. Here, it seemed, was a new way of demonstrating musical coherence in non-monotonal music. But the implications of the sonorità concept are more radical than this. As I have already noted, the sonorità de-privileges the bass of a musical texture, using melodic pitches (not tonics) as the principal agent of coherence. It thus represents a departure from all bass-oriented theories of tonality. One task of opera analysis in the coming decades will be to evaluate the applicability and scope of the sonorità concept.

[14] Unlike the early Wagner analyst Alfred Lorenz, Verdi scholars have rarely invoked Hugo Riemann. But the Riemann revival of recent years reminds us that Riemann, too, argued against privileging the bass; he railed ceaselessly against what he called “thoroughbass” theories, among which he included Rameau’s fundamental bass and Gottfried Weber’s Roman numerals. Riemann’s function theory has no fixed points of reference except the tonic and its two dominants; his theory of Schritte and Wechseln lacks even these. A theory of nineteenth-century music based on common-tone relations—Kopp’s term “common-tone tonality” describes it nicely—has much in common with Riemannian and especially neo-Riemannian theories. It also has obvious points of contact with theories of twentieth-century “centric” music. Igor Stravinsky, for example, pointed to the note A as the “axis of sound” in his Sérénade en la (1925), regardless of whether A is harmonized by chords of F major, A minor, or (later) A major. In other words, Stravinsky’s la is a sonorità.⁽²⁷⁾

[15] We are left with multiple theoretical traditions, German and non-German, with no clear relation to each other. In what follows, I begin to bring some of these traditions together for the purpose of understanding Italian operatic music between 1810 and 1860. I proceed more or less chronologically, beginning with Rossini and ending with a famous excerpt from Verdi’s Il trovatore. As we shall see, Rossini embeds chromatic mediants and tonal pairing within a mostly traditional tonal framework. Composers from Bellini onward increasingly disrupt this framework, moving into realms beyond monotonality. To explain these later practices (which begin no later than 1825), we will need to employ theoretical approaches not typically applied to music from the first half of the nineteenth century.

IV.

[16] The frequent use of chromatic mediants is a Rossinian fingerprint from the earliest years of his career. What made the cabaletta “Di tanti palpiti” from Tancredi (1813) wildly popular for half a century was only partly its jaunty melody; it was also the quietly thrilling move from a C-major chord (representing a half cadence in F major) to a new phrase in A-flat major, with nothing to foreshadow it save a crescendo on scale-degree $\hat{2}$, which becomes the leading tone to A. (See Example 1.) The characteristically Rossinian move of a major third downward, from one major triad to another, was far from unknown in eighteenth-century music, where it usually involved a direct move from III, approached as V of the relative minor, to a major tonic. In an article first published in 1957, Jan LaRue called this phenomenon “bifocal tonality,” which he described as a combination of a major key with its relative (not parallel) minor.⁽²⁸⁾ The presence of III–I progressions in the music of J. S. Bach, Haydn, Mozart, and Beethoven has been widely discussed; in his Harmonielehre (1906), Schenker describes III–I as a means of tonicization, an alternative to V–I.⁽²⁹⁾ Rossini’s move in Tancredi is more unconventional, because it involves not a return to the tonic but a leap from an active dominant into a distant key.

[17] Rossini’s progression can easily be described using any of the available neo-Riemannian theories. In the best-known version, based on the work of David Lewin and others, Rossini’s move from a C-major to an A-major triad represents the double transformation PL; the change of key, F major to A-flat major (a more abstract relation), represents PR. Using Kopp’s positional terms—not his transformations—Rossini moves from V to UFM (the Upper Flat Mediant) of his main tonic, F major; in relation to A-flat major, the move is from USM (the Upper Sharp Mediant) to I. In terms of Daniel Harrison’s theory, there is a dominant discharge from G to A (leading tone to tonic) and a subdominant discharge from E (representing F, the flatted sixth degree) to E.⁽³⁰⁾ Readers who are so inclined may translate my remaining analyses into the desired dialect of neo-Riemannian.

[18] Although Rossini’s music is not as stereotyped as some believe, his most effective ideas often turned into predictable mannerisms. In 1818, less than two years after Rossini’s music was first performed in Vienna, a critic there wrote: “[Rossini] generally precedes the stretta of the first finale with a very slow movement in a dark but theatrically effective A-flat major, whereupon the noisy stretta follows in a bright C major.”⁽³¹⁾ The critic may have been describing the first-act finale of Il barbiere di Siviglia (1816), which follows precisely this pattern. He might have added that if a Rossini finale ends in C major, it also begins there; a slow movement in A-flat major would thus be heard as VI in relation to the finale’s overall tonic. In 1816, Rossini—a classicist in many respects—still believed in unity of key over the course of an operatic number, whether that number was an introduzione, a double aria, a four-movement duet, or a multi-movement finale. The slow movement of any of these tends to be set by Rossini in a chromatic mediant key, either III, III, or VI. (Rossini did not favor VI for this purpose.)

[19] The first-act finale of Semiramide (1823) is also in C major, but its slow movement—known, in a finale, as the pezzo concertato—goes a step further than usual. Rossini begins this concertato in A-flat minor, although he uses a key signature of only four flats; he resolves this extremely distant key to the usual A-flat major midway through the movement. And yes, the final movement is a “noisy stretta . . . in a bright C major.”

[20] Especially (but not only) in comic operas, Rossini found sudden shifts to chromatic mediants useful to depict outbursts of surprise, frustration, or rage on the parts of his characters. The flat chromatic mediants, III and VI, are heavily favored for this purpose. In Barbiere a surprising number of these chromatic outbursts involve the triad of E-flat major, acting either as III in C major, as VI in G major, or in one case as II in D major (in Basilio’s aria “La calunnia”). The focus on E-flat major is especially striking in Act 1, in which all numbers but one are in sharp or neutral (“white-note”) keys. The exception is the act’s penultimate number, Bartolo’s aria “A un dottor della mia sorte,” which is in E-flat major. Once the finale begins, Bartolo’s key is recontextualized as III of C major. The only other number in Act 1 with a key signature of more than two sharps or flats is Rosina’s cavatina “Una voce poco fa,” which is in E major—VI of the act-opening G major, III of the finale’s C major.⁽³²⁾

[21] In Barbiere, the chromatic outbursts that I have described generally explain themselves retrospectively; their entrances are disruptive, but their exits are well-mannered. Even the noisiest VI, for example, usually resolves to V (although there is a complete major-third cycle I–VI–III–I in Figaro’s “Largo al factotum”). Such conventional contextualizations of chromatic mediants occur less consistently where the slow movement of a multi-movement number is concerned. Usually a modulating transition will lead fairly smoothly into the slow movement, but the movement itself is tonally closed, ending with repeated PACs in its assigned key. The following tempo di mezzo (a transitional movement) may begin directly with the governing tonic, so that the underlying chromatic third-relation is heard directly; or the tempo di mezzo may prolong the home dominant, so that the slow movement’s chromatic mediant is heard to progress to V; or the tempo di mezzo may begin in some foreign key and work its way back to the tonic. When there is direct motion from the chromatic mediant to the tonic, the effect may be compared to that in some of Haydn’s late instrumental works in which a slow movement in some distant, flat-side key is followed by a minuet or finale in the tonic.⁽³³⁾

[22] Some of Rossini’s most adventurous harmonic writing is found in his last opera, Guillaume Tell (1829). A significant portion of Tell—the famous Overture, the five-movement Introduction, the four-movement finale to Act 1, and much of Act 3—is dominated by the pairing of two tonics, E and G.⁽³⁴⁾ G appears almost exclusively as G major; E vacillates between its minor and major modes. (C major appears often as an intermediary key.) In effect, E minor is surrounded by its relative and parallel major keys—recalling, perhaps, Gottfried Weber’s well-known chart of key relations (Example 2), in which each minor key is surrounded in precisely this fashion on the chart’s horizontal axis.⁽³⁵⁾ We know that Robert Schumann was familiar with Weber’s chart; ⁽³⁶⁾ whether Rossini knew it is uncertain.

[23] Tell is Rossini’s most Romantic opera not only aesthetically but compositionally. Employing both relative and parallel keys is nothing new, but Rossini employs them in a decidedly post-classical way. By “post-classical” I refer primarily to three phenomena: (1) the frequent use of direct modulation by common tone, foregoing dominant preparation of the new key; (2) use of the relative major outside the I–III–V bass arpeggiation described by Schenker as normative for the minor mode; (3) in major, using III–I progressions that are not hierarchically subordinate to some larger fifth-relation.⁽³⁷⁾

[24] In Tell, no fewer than three numbers—the finales of Acts 2, 3, and 4—violate Rossini’s usual practice of beginning and ending a number (excluding any introductory recitative) in the same key, perhaps with a change from minor to parallel major.⁽³⁸⁾ However, all numbers that feature the E–G pairing are tonally closed; that is, they begin and end with the same tonic. The Overture, in four short movements, traces the key scheme e/E–e–G–E. Despite its nontraditional form, the Overture’s tonal arc is sufficiently traditional to be represented in a Schenkerian graph, which also reveals some (though not all) of the music’s unusual features. Example 3 offers such a graph. It shows that G major is repeatedly reached from V of e/E, a major triad in either mode. The famous stretta (Allegro vivace, E major, 2/4) is announced by a trumpet call on B4, B being the one tone common to the tonic triads of all three keys. Within the stretta, E major alternates several times with its relative minor, c. Moves from E to c are made directly, from one tonic to the other, with emphasis on the common tone G4. The quiet middle section (measures 316–43) consists of twelve measures in C minor followed by sixteen measures in E major, linked by the chord of G minor—minor dominant of c, diatonic mediant of E—and the melodic tone G4.⁽³⁹⁾

[25] Example 4 gives most of the Overture’s first movement in a piano reduction. The head-motive of the main theme, in E major, is harmonized differently on each of its four appearances (measures 17, 21, 28, and 32). The last two of these harmonizations include the pitch-class E/F, which was introduced in a highly marked way in measures 12–13, where it was dramatized by the unusual leap of a diminished fifth—unusual, that is, in the progression from a Neapolitan sixth to a dominant. The F-major chord returns at the upbeat to measure 34, now not as Neapolitan but as subdominant of a C major that returns immediately to E major. In the ensuing cadential progression, a conventional trill on scale-degree $\hat{2}$ (F4) is accompanied by some rather unconventional harmonies: in addition to V of E, we hear dominant seventh chords on G and D, in each of which the trilled F is heard as an unresolved tendency tone, either leading tone or seventh.⁽⁴⁰⁾ Within measures 37–40 F is heard variously as root, third, fifth, and seventh. Rossini avoids only sounding F as a minor ninth above a bass of E/F; doing so would have completed a minor-third cycle in the bass. In short, the Overture’s first movement exhibits some features associated with post-classical tonality, including common-tone shifts and unresolved dissonant chords.

[26] Example 5 sketches the orchestral transition between movements 2 and 3 of the Introduction (the opening number of Act 1). The Introduction’s five movements trace the key scheme G–C–e–E–G, so the overall modulation in our example is from C major to E minor. The decisive move is the first, from C major to E major, which is carried out by first isolating the common tone E4 in the solo horn (the stage directions read “Ici l’on entend le ranz des vaches [here one hears the ranz des vaches]”). Motion to the implied G-major harmony is from V of E, another III–I progression in a major key; the common tone B (heard in two registers) forms the link. As my sketch shows, G major is embedded within the transition in a rather classical way, forming part of a dominant prolongation in E minor. G returns to V of E through a harmonic sequence and a descending fourth-progression in the upper voice, the latter outlining the interval from root to fifth of the V harmony. Repeated soundings of B3 and B4 lead to the third movement, an Allegro vivace in E minor. A similar but shorter transition links movements 4 and 5, again emphasizing B4 as common tone between the tonic triads of E major and G major.

[27] To summarize: Rossini was an early adopter of chromatic mediants, which he used chiefly for purposes of dramatic contrast. It is in this area that he most clearly influenced Austro-German composers, beginning with the first Viennese performances of his operas in 1816.⁽⁴¹⁾ Rossini was more conservative about directional tonality and tonal pairing, Guillaume Tell being a conspicuous exception. His frequent use of common-tone modulation, in which the common tone is often isolated or otherwise highlighted, adumbrates the use of the sonorità as a unifying device. His chromaticism, which I have only touched upon here, was sufficiently advanced for Fétis to count him among the pioneers of the genre omnitonique. Tonal closure within an operatic number was a binding requirement for the young Rossini and remained the norm thereafter. Nevertheless, the enormous size of some numbers, their division into several movements, and the freedom of modulation within movements made the move away from monotonality relatively easy, for later composers if not for Rossini himself.

V.

[28] Donizetti’s operatic career began in 1818, Bellini’s in 1825. I am unfamiliar with most of Donizetti’s early operas, but there are numbers in La zingara (1822) that begin and end in different keys.⁽⁴²⁾ Interestingly, this opera was a favorite of Bellini’s in his student years. While both composers felt free to begin and end a number in different keys, Bellini used tonal pairing—which involves repeated motion between two tonics—remarkably early and in an especially prominent way. He also uses fixed melodic pitches in ways that seem to have little precedent in Rossini.

[29] In Bellini’s first opera, Adelson e Salvini (1825), the duet of Salvini and Bonifacio begins in F minor and ends in C major. The orchestral prelude begins with a unison C repeated many times, and C acts as a dominant pedal almost continuously throughout the prelude. The duet’s initial F-minor strain ends with an emphatic half cadence, whereupon the unison C returns (again repeated many times) and leads to A-flat major in a III–I progression. After this, a modulatory transition leads to another half cadence in F minor (the tonic appears only in passing), to be followed once again by the repeated unison C and the A-flat-major strain. A lengthy dominant pedal of C minor then leads to the C-major cabaletta. It could easily be argued that the duet is forged into a musical unit by the note C, which acts successively as dominant of F minor, third of A-flat major, and tonic of C major. This C does not quite fit Petrobelli’s definition of a sonorità because it is emphasized primarily in the orchestra, not in the vocal lines.

[30] In The Operas of Verdi, Julian Budden cites Gualtiero’s cavatina, or entrance aria, in Il pirata (1827) as an early example of a number that begins and ends in different keys.⁽⁴³⁾ Budden seems to be referring principally to the first movement, or cantabile, of Gualtiero’s aria, but the path of directional tonality that defines the cantabile—G minor to B-flat major—also defines the aria as a whole. It also matches the tonal trajectory of Act 1 in the opera’s first printed score, a vocal score published in 1828.⁽⁴⁴⁾ The pairing of G (mostly minor) and B (mostly major) also pervades Gualtiero’s scena (recitative), which is not counted as part of the aria proper.

[31] Example 6 gives a simple diagram for the aria as a whole; Example 7 is an excerpt. The scena vacillates between B-flat major and G minor. The aria’s opening ritornello begins solidly in G minor. It modulates to D minor just as the tenor enters, singing the final word of his scena, “ascolta” (“listen”).⁽⁴⁵⁾ And listen we do, for the D-minor cadence is followed immediately by a fortissimo dominant seventh of B-flat major, which is followed in turn (after a brief silence) by a quiet PAC in G minor. Providing dominant preparation for one key but delivering its relative becomes an Italian trademark for decades to come.⁽⁴⁶⁾

[32] The main body of Gualtiero’s cantabile repeats the ritornello as far as the internal cadence in D minor. After the fermata, we hear a completely new strain in B-flat major, entirely over a tonic pedal; with this the first strophe will end. After a modulating middle section, the opening strophe is heard a second time. The following tempo di mezzo returns to the aria’s original tonic, G, but this time it is G major. After a spell in related sharp keys (E minor and B minor), Bellini modulates by way of G minor back to B-flat major for the cabaletta. The latter is firmly in B-flat major, with internal cadences in G minor.

[33] The aria’s tonal plot could hardly be clearer. Two years before Rossini’s Guillaume Tell, Bellini pairs two tonics, G and B, within a trio of keys in which G minor is flanked by its relative and parallel majors. Unlike Rossini, Bellini allows free motion between his tonics, with no obligation of ending where he began, either within a single movement or over the course of a multi-movement aria.

[34] Bellini’s three tonic triads share a single pitch class, D. Not surprisingly, D4 is the focal pitch in Gualtiero’s vocal line. D4 is more than a Schenkerian Kopfton (although it is that, too); it is more like a reciting tone. Furthermore, Bellini repeats an unordered group of three pitches, F4–E4–D4, within both the G-minor and the B-flat-major portions of the aria. In example 7 I have bracketed this three-note cell on many of its appearances. Toward the end of the scena (measures 41–45), the cell’s F is pointedly cross-related to the dominant’s F. Here F is an appoggiatura to an upper-neighbor E, as unstable a tone as can be imagined within a largely diatonic passage. During the orchestral ritornello, the oboe repeats the three-note cell obsessively—much as Gualtiero, the pirate of the opera’s title, has dwelt on the mental image of his beloved Imogene during his years at sea. F remains an upper neighbor to an upper neighbor. But when it comes time actually to represent Imogene’s image—to fix it within his mind—Gualtiero fixes the three-note cell within a B-flat-major harmony, so that F becomes consonant. The psychological aptness of the musical transformation is clear.

[35] The cantabile is in the usual two-strophe form. In its brief coda, the three-note cell is sung slowly over a tonic pedal in B-flat (Example 8), imprinting it still more firmly. The cell disappears from the tempo di mezzo but is heard repeatedly at the beginning of the cabaletta (Example 9)—first ascending, consonant, and over a major triad, as when Gualtiero conjured Imogene’s image in the cantabile; then descending, dissonant, and over a minor triad, as in the earlier G-minor music. Elementary Schenkerian analysis of the opening phrase reveals the presence of the cell just beneath the musical surface (Example 10), an instance of so-called hidden repetition. Curiously, this cabaletta is a transposed version of the cabaletta from the Salvini/Bonifacio duet in Adelson e Salvini. Recycling pieces from little-known or failed operas was common practice for Rossini and Bellini, so its use here is not surprising. What is impressive is that Bellini realized just how well his old wine suits its new bottle.

[36] Bellini’s recurring use of a three-note scale fragment common to two keys suggests that Petrobelli’s sonorità concept may be unnecessarily restrictive in its focus on a single pitch. If Gualtiero has a single sonorità, it is surely D4. But recognizing the three-note cell as a unifying entity seems a better analytical strategy for this aria.

[37] Directional tonality and tonal pairing remain exceptional in Il pirata and in Bellini’s next opera, La straniera (1829). When they do occur, therefore, they are especially striking. The rondò-finale (concluding aria) of La straniera includes a cantabile in B-flat major, a widely modulating tempo di mezzo, and a cabeletta that moves from a B-flat-minor beginning to a D-flat-major ending; the contrasting interlude between the cabaletta’s two strophes provides dominant preparation for B-flat minor, as did the end of the tempo di mezzo. One notes a similarity between the cabaletta’s key scheme and that of Chopin’s Scherzo op. 31 (1837); although Chopin surely knew La straniera, the similarity may be coincidental. Owing to the presence of the B-flat-major cantabile, it is difficult and probably nonsensical to hear the cabaletta’s opening B-flat minor merely as VI of D, as Schenker argued should be done in Chopin’s op. 31.⁽⁴⁷⁾ Unlike Gualtiero’s cavatina from Il pirata, there is no obvious sonorità or recurring pitch-cell in Alaide’s rondò.

VI.

[38] In Italian operas of the 1830s, tonal unity within an operatic number is at best a weak stylistic expectation, violated as often as it is fulfilled. By 1850 a tonally unified number sounds faintly archaic. The forms inherited from Rossini become a principal means of orientation for listeners, without stable tonal props to support them. Cantabiles and cabalettas often remain tonally closed, but larger numbers (of which these are components) do not.

[39] It is impossible to generalize about tonal architecture in Verdi’s operas. Some operas seem more tightly constructed than others, but the means on which Verdi relies for tonal association and contrast are not consistent from one opera to the next.⁽⁴⁸⁾ Not surprisingly, operas that were never substantially revised once completed, such as Rigoletto and Un ballo in maschera, tend to have clearer structures than those, such as Don Carlos, that went through multiple versions. Some operas, such as Macbeth (1847, rev. 1865) and Simon Boccanegra (1857, rev. 1881), rely on major-third cycles and their associated hexatonic systems for their principal keys, each of which may have clear dramatic associations.⁽⁴⁹⁾ Each Verdi opera inhabits its own world and must be evaluated independently.

[40] Il trovatore has attracted more analytical comment than any middle-period Verdi opera except, perhaps, Rigoletto. The most obvious thread running through the opera, especially for listeners with absolute pitch, is the opposition between mildly sharp-side or neutral keys (C major, A minor, E minor, and G major) and keys with four or more flats, especially the trio of keys F minor–A-flat major–A-flat minor (I list them in an order opposite to Weber’s so that the most extreme key appears last). Several commentators have convincingly associated the former set of keys with the gypsy Azucena and the gypsies in general; the latter, flatter set is associated with the high-born Leonora.⁽⁵⁰⁾ Aurally, the gypsies inhabit a realm of light, Leonora one of darkness.

[41] The opening number of Act 2 was entitled Coro di Zingari e canzone (“Chorus of Gypsies and song”) by Verdi.⁽⁵¹⁾ It is popularly known as the “Anvil Chorus,” although Azucena’s canzone “Stride la vampa” is often erroneously regarded as a separate number. By “Anvil Chorus” I designate the entire number as Verdi conceived it. In this number we again encounter the kinds of fixed-pitch references that were such a striking feature of the Bellini aria analyzed earlier. These references are so little dependent on their harmonization that they recall pre-tonal settings of cantus firmi—for example, Monteverdi’s settings of psalm tones in his 1610 Vespers.⁽⁵²⁾

[42] I will claim that Verdi’s “Anvil Chorus” is “in” E in much the same sense that Stravinsky’s Sérénade en la is “in” A. The piece begins in E minor and ends in C major, with E prominently featured at both beginning and end. E gradually emerges as a sonorità for the chorus, acting as a reciting tone first over V of A minor, then over I of C major (another III–I progression). The chorus’s final cadence occurs on the melodic notes G–F–E, doubled in octaves.

[43] Azucena’s two-strophe canzone acts as a middle section within the “Anvil Chorus” as a whole. As Petrobelli has pointed out, Azucena’s vocal line features a sonorità of B4, sounding here as the fifth degree in E minor; her line descends by step within a conventional parallel period (scale degrees $\hat{5}$–$\hat{2}$ in the antecedent, $\hat{5}$–$\hat{1}$ in the consequent). In the second half of her strophe, B4 is increasingly overtaken by E5, which is harmonized in multiple ways within tonicizations of C major, A minor, and E minor. Azucena’s final cadence, G5–F5–E5 ($\hat{3}$–$\hat{2}$–$\hat{1}$ in E minor), conspicuously echoes the chorus’s earlier cadence, G5–F5–E5 ($\hat{5}$–$\hat{4}$–$\hat{3}$ in C major), especially since the final three pitches are approached similarly, are heard in similar rhythms, and receive similar orchestral punctuation (see Example 11). Following the canzone there is an abbreviated reprise of the chorus, followed by a quiet orchestral postlude. Only in the last two measures of the postlude is the melodic E relinquished in favor of the local tonic, the E–C third forming a link to the following number (“Racconto d’Azucena”).

[44] Example 12 diagrams the melody of the “Anvil Chorus”; readers should also consult a score, either orchestral or piano-vocal. In the following discussion, I analyze Verdi’s melody using terms borrowed from modal and neo-modal theory.

[45] The principal ambitus of the melody lies between E4 and E5, solidly within Azucena’s mezzo-soprano range. The opening orchestral melody begins lower than this, however: starting on E4, it repeatedly touches the fifth degree both below and above, moving within the ambitus B3–B4, as though in a plagal mode (E Hypoaeolian). Once the key changes to G major, still within the orchestral prelude, the ambitus shifts upward, with the violins “reciting” on B4. When journeying beyond their reciting tone, the violins touch first on E5, then G5, before cadencing on G4; a G4–G5 octave is briefly outlined. The unison chorus then enters on G4 and moves upward by step to E5. As the orchestra settles onto V of A minor, E is heavily emphasized in both chorus and orchestra. The “anvil” theme features E as a reciting tone and decorates it with F, its semitone upper neighbor. Nevertheless, the melody moves mostly downward from E, traversing the octave E5–E4 (here I privilege the women’s register within the unison chorus) as if outlining the authentic mode on E (E Phrygian). The final climax is on G5, the highest sustained pitch in the entire number, followed by the G–F–E descent mentioned earlier. E has emerged as the melodic final, and the authentic octave, E4–E5, as the principal ambitus. G5 lies outside this ambitus but is used as a climax note; in Joachim Burmeister’s terms, it represents the figure hyperbole.⁽⁵³⁾ E5 has been featured prominently in four different keys: E minor, G major, A minor, and C major. The outlining of the modal fourth B4–E5 over a G-major harmony (measures 13–14) is an especially noteworthy instance of the melody’s independence from its accompaniment.⁽⁵⁴⁾

[46] The opening of the canzone remains within the E4–E5 ambitus while emphasizing B4 as a reciting tone. A secondary sonorità of E5 is established, as we have seen, and soon becomes the principal focus; it is again exceeded by the cadential figure, this time G5–F5–E5. I have already noted the musical rhyme between the end of the chorus and that of the canzone. The similarity of ending formulas emphasizes the change from the neutral system (with F) to the one-sharp system (with F). In neo-modal terms, the change could be described as one from E Phrygian to E Aeolian and back again (in the reprise of the chorus), following the Hypoaeolian orchestral prelude.

[47] Square brackets in example 12 demonstrate melodic correspondences within the “Anvil Chorus.” The legend at the bottom of the example labels individual motives, all of them scale fragments ranging from a second to a sixth in length. The number assigned to each motive indicates the size of its boundary interval: 2 for a second, 3 for a third, etc. Motives with the same number and letter share the same pitch level; for example, motive 3B outlines the third from B4 to G4 in either direction. (All motives are shown descending, but this ordering is arbitrary.) The restriction to stepwise motives exhibits a certain Schenkerian bias, but it permits me to show stepwise motives that lie a little beneath the surface, such as the instance of motive 3B very near the beginning (measures 2–3). It is interesting that, within the orchestral prelude, motive 3B appears three times and in two different keys, E minor and G major. A disadvantage of the limitation to stepwise motives is that some quite audible correspondences do not appear, including one between measure 1 and measure 207, where the melody shares the descending fourth E4–B3, representing the lower fourth of a plagal ambitus.

VII.

[48] Especially in his later years, Verdi argued that Italian composers should focus their study on the Italian musical inheritance, not that of the Germans. The Germans do well to study Bach, he said, but Italians should learn from their Italian predecessors, stretching back to Palestrina. Verdi put this pedagogical philosophy into practice as early as the 1840s in his teaching of Emanuele Muzio, his longtime amanuensis.⁽⁵⁵⁾ I do not know how much early music Verdi knew, but it is remarkable how well the “Anvil Chorus” responds to theoretical concepts that predate major-minor tonality. Perhaps Verdi’s technique in Il trovatore seems innovative to us because, in some ways, it is so old.

[49] Bellini’s “Nel furor della tempesta” (Gualtiero’s cavatina) and Verdi’s “Anvil Chorus” have much in common. They do not so much explode the concept of bass-oriented tonal unity as they ignore it, and they do so in an overwhelmingly diatonic way. The musical language of both composers becomes more chromatic in later works (e.g., Bellini’s Beatrice di Tenda and Verdi’s Otello), but the excerpts examined here exhibit only the mildest chromaticism, thus contravening the idea (which apparently began with Fétis) that the dissolution of classical tonality is causally bound to increasing chromaticism and enharmonicism. Bailey, following Schoenberg, claims that Wagner eliminated the duality of major and minor and substituted, for twenty-four keys, twelve “chromatic keys” based on the tonic alone.⁽⁵⁶⁾ Without addressing whether this is an accurate description of Wagner’s practice overall (it surely is for parts of Tristan and Parsifal), it is untrue of the Italian composers who pioneered the use of tonal pairing. For them, what dissolved was less the distinction between parallel major and minor than that between relative major and minor. The parallel relation remained available, to be sure, but it was a supra-tonal diatonicism that Italian composers called upon when they wanted to combine tonal mobility with simplicity of melodic surface, as they often did. Sonorità and diatonic system became more important as unifying factors than the choice of final note in the bass (tonic).

[50] Precedents for nineteenth-century Italian diatonicism may be found in the eighteenth century, mostly outside of Italy. Rameau wrote at length, in his Code de musique pratique (1760), on the possibility of harmonizing the same melodic “cadences” in a major key and its relative minor, and on the interlacing (entrelacement) of keys that may result. In the opera house, eighteenth-century audiences could hear Rameau’s interlacing of relative keys in the operatic romance. Daniel Heartz has traced the early history of this genre, which began in mid-century France.⁽⁵⁷⁾ As Heartz notes, many romances shifted from a minor key to its parallel major, but the self-conscious archaism of the genre sometimes led composers to write modal-sounding melodies in which the boundary between relative keys was blurred.⁽⁵⁸⁾

[51] The most remarkable of eighteenth-century romances is surely “Im Mohrenland” from Mozart’s Die Entführung aus dem Serail (1782). The shifting keys of this romance (example 13), which dissolves into spoken dialogue and thus has no fixed ending,⁽⁵⁹⁾ can perhaps best be understood within a referential system of two sharps and a melodic final of F–F Phrygian, in other words. If our analysis addresses system, final, and ambitus (Pedrillo’s is F authentic), we are speaking in terms of Harold Powers’ tonal types, not those of major-minor tonality.⁽⁶⁰⁾ Since we are dealing with a musical texture in which there is one leading melody rather than a polyphonic web, there is no conflict between mode and tonal type.

[52] A recent study by Balázs Mikusi explores the interpenetration of relative keys in Felix Mendelssohn’s “Scottish” works.⁽⁶¹⁾ Mikusi notes that Mendelssohn also blends relative keys in other pieces in which an exotic flavor is intended, including the first of his “Venetian Gondola Songs” (G minor, no. 6 of the Lieder ohne Worte op. 19b, composed in 1830). The seven-measure introduction to the Gondola Song touches strongly on a tonicized B-flat major, albeit in a syntactically classical way, as III in the bass arpeggiation I–III–V–I. The four-note melody, representing the cry of the gondolier (obviously a tenor), is D4–F4–E4–D4, recalling Gualtiero’s pitch-cell in Il pirata (1827).

[53] All of the compositional examples of entrelacement that we have just considered were motivated by exoticism, a desire to portray the Other, whether that Other resided at a distance of time (the neo-medieval French romance) or of space (Mozart’s “Moorish land,” Mendelssohn’s Scotland) from the composer’s here and now. To a lesser degree, Othering may be detected in our two principal nineteenth-century examples, Bellini’s Gualtiero (an outlaw, though of noble birth) and Verdi’s Azucena (a gypsy and an outlaw). I chose these examples because they portray nineteenth-century Italian diatonicism, including its disregard for tonal unity, in an especially pure and striking form. On the other hand, what was exotic for Mozart and Mendelssohn quickly became routine for Bellini, Donizetti, and Verdi. Most Italian examples that one could cite betray no obvious exoticism. Common-tone tonality is a normal part of the Italian musical language from the late 1820s onward. As Kaminsky demonstrated in his study of Schumann’s piano cycles, it became normal for Schumann by the mid-1830s. The position of Schubert in this development—not only the degree of Rossini’s influence on Schubert, but the degree of Schubert’s influence on his immediate successors—is less clear and requires further research.

VIII.

[54] The notion that music of the eighteenth and nineteenth centuries represents a “common practice” is a myth whose burial is long overdue. We have seen that classical tonality (as I choose to call it for want of a better term) does not always apply even to Mozart. Vestiges of pre-tonal practice survive to the end of the eighteenth century; post-classical practice begins soon after. An historical link between the two seems probable, especially in Catholic parts of Europe, where plainchant continued to be sung in variously modified forms.⁽⁶²⁾ A later stage of this development is represented by the neo-modality of late-nineteenth-century French music (linked to the Gregorian revival) and, later still, by the neoclassicism of the early twentieth century. These developments all relate to what might be called the diatonic genus of common-tone tonality. The chromatic genus, which is the principal focus of neo-Riemannian theory, deserves separate historical consideration, which would again include Rossini and Verdi but also Meyerbeer and, of course, Wagner.

[55] Music theorists do not generally write music histories, but their perspective can benefit those who do. Nineteenth-century Italian opera is a worthy object of analytical study for its own sake. Such study can also teach us much about the history of compositional technique, and the contribution of composers both Italian and non-Italian to a European musical language that was undergoing rapid change, especially in the century’s middle decades. To see these developments clearly, however, we need to unlearn the unitary historical narratives imposed by German Romanticism and its offspring, Viennese Modernism. Theirs is not the only story that needs to be told, and their tendency to claim exclusive possession of aesthetic truth should be resisted. The tales of Hoffmann, Schumann, and Marx in the nineteenth century, and of Schenker and Schoenberg in the twentieth, were designed (in times of great political stress) first to establish German hegemony in music, then to preserve or regain it. Twenty-first-century North Americans need not be bewitched by their histories, their nationalism, or their aesthetics.

William Rothstein
Queens College and The Graduate Center, City University of New York
365 Fifth Avenue
New York, NY 10016
wrothstein@gc.cuny.edu

Footnotes

1.   Carl von Clausewitz, Vom Kriege (Berlin, 1832), “Nachricht.”
Return to text

2. For a recent specimen see Taruskin, “The Musical Mystique: Defending classical music against its devotees,” in The New Republic 237, No. 4,823 (October 22, 2007): 34–45.
Return to text

3. Although German Romanticism included aesthetic stances other than organicism—notably the cult of the fragmentary pursued by Robert Schumann in the first half of his career—it seems undeniable that organicism is the dominant strain where music is concerned.
Return to text

4. The term “Carpenter narrative” (after the late Patricia Carpenter) was coined by M. Murray Dineen in his recent writings on Schoenbergian analysis. See his “Tonal Problem, Carpenter Narrative, and Carpenter Motive in Schubert’s Impromptu, Op. 90, No. 3,” Theory and Practice 30 (2005): 97–120.
Return to text

5. See Celia Applegate and Pamela Potter, eds., Music and German National Identity (University of Chicago Press, 2002).
Return to text

6. The notion of Italian opera as counterculture derives in part from Carl Dahlhaus, Nineteenth-Century Music, trans. J. Bradford Robinson (Berkeley: University of California Press, 1989).
Return to text

7. Robert Bailey, “The Structure of the Ring and Its Evolution,” 19th-Century Music 1, no. 1 (1977): 48–61; see also “An Analytical Study of the Sketches and Drafts,” in Wagner, Prelude and Transfiguration from Tristan und Isolde, ed. Bailey (New York: Norton, 1985), 113–46. Some concepts were introduced, without their now-familiar names, in Bailey’s Ph.D. dissertation, “The Genesis of Tristan und Isolde and a Study of the Sketches and Drafts for the First Act” (Princeton University, 1969).
Return to text

8. William Kinderman and Harald Krebs, eds., The Second Practice of Nineteenth-Century Tonality (Lincoln: University of Nebraska Press, 1996).
Return to text

9. See, e.g., “Alternatives to Monotonality in Early Nineteenth-Century Music,” Journal of Music Theory 25, no. 1 (1981): 1–16; “Some early examples of tonal pairing: Schubert’s ‘Meeres Stille’ and ‘Der Wanderer’,” in The Second Practice, 17–33.
Return to text

10. Kinderman, “Directional tonality in Chopin,” in Jim Samson, ed., Chopin Studies (Cambridge: Cambridge University Press, 1988), 59–75; Korsyn, “Directional Tonality and Intertextuality: Brahms’s Quintet op. 88 and Chopin’s Ballade op. 38,” in The Second Practice, 45–83.
Return to text

11. Kaminsky, “Principles of Formal Structure in Schumann’s Early Piano Cycles,” Music Theory Spectrum 11, no. 2 (1989): 207–25.
Return to text

12. On the canonization of Chopin in German musical culture see Samson, “Chopin reception: theory, history, analysis,” in John Rink and Jim Samson, eds., Chopin Studies 2 (Cambridge: Cambridge University Press, 1994), 1–17.
Return to text

13. The Second Practice, 12.
Return to text

14. These major-third relations were recognized as independent relations of a tonic as early as Rameau’s Génération harmonique (1737). Riemann gave them their own function symbols in the sixth edition of Handbuch der Harmonielehre (1917).
Return to text

15. Schumann railed equally against Italian opera, Meyerbeer, and the cult of the virtuoso. Both Schumann and Wagner became more chauvinistic over the course of their careers. That Schenker held a low opinion of opera in general, not only Italian opera, is shown by his recently published Über den Niedergang des Kompositionskunst (“The Decline of the Art of Composition”), Music Analysis 24/1–2 (2005): 3–232. Interestingly, both Hanslick and Schenker modified their critical view somewhat where Verdi’s last operas were concerned; Hanslick in particular saw these as an approach toward German musical style.
Return to text

16. A strain of Italophilia (at least Verdi-philia) survived among some American composers, notably Roger Sessions and Donald Martino, both of whom enjoyed close friendships with Luigi Dallapiccola, an ardent Verdian. This seems to have had few consequences for their theoretical teaching, however. The list of “References to Musical Works” in Sessions’ Harmonic Practice (1951), for example, includes a bit more Berlioz than one would find today, but otherwise it is very much the usual German repertoire. Italian opera is represented by a single chord from Rigoletto.
Return to text

17. See Angela Pachovsky, “Zur Frage der Rossinismen bei Schubert,” and Xaver Hascher, “Quelques paradigmes harmoniques dans la forme sonate chez Schubert,” both in Dietrich Berke et al., eds., Franz Schubert: Werk und Rezeption, II: Bühnen- und Orchesterwerke, Kammer- und Klaviermusik (Kassel: Bärenreiter, 2000), 35–55 and 151–61.
Return to text

18. But see Friedrich Lippmann, “Wagner und Italien,” in Lippmann, ed., Colloquium “Verdi–Wagner” Rom 1969 (Analecta musicologica 11; Vienna: Böhlau, 1972), 200–249.
Return to text

19. Carolyn Abbate and Roger Parker, eds., Analyzing Opera: Verdi and Wagner (Berkeley: University of California Press, 1989).
Return to text

20. See the collection cited in note 17.
Return to text

21. I focus on Verdi scholarship because it has dug deeper, analytically, than have writings on Rossini, Bellini, or Donizetti.
Return to text

22. There have been individual articles on Verdi’s music by two distinguished theorists on the American scene, Matthew Brown (on Un ballo in maschera) and John Roeder (on the Requiem). Brown may be unique among English-language theorists in having published important articles on both Verdi and Wagner. Richard Cohn has occasionally cited passages by Verdi in his writings on neo-Riemannian theory.
Return to text

23. Steven Huebner has applied William Caplin’s Formenlehre to the “lyric form” of mid-century Italian opera; see “Lyric Form in Ottocento Opera,” Journal of the Royal Musical Association 117, no. 1 (1992): 123–47.
Return to text

24. Petrobelli, “Per un’esegesi della struttura drammatica del Trovatore,” in Atti del terzo congresso internazionale di studi verdiani (Parma, 1974); trans. William Drabkin as “Toward an Explanation of the Dramatic Structure of Il trovatore,” Music Analysis 1, no. 2 (1982): 129–41; rept. in Petrobelli, Music in the Theater: Essays on Verdi and Other Composers (Princeton University Press, 1994), 100–112. Petrobelli was surely familiar with Abramo Basevi’s Studio sulle opere di Giuseppe Verdi (Florence, 1859), in which Basevi suggested that the beginning of Azucena’s canzone “Stride la vampa” could be harmonized almost equally well in E minor or G major. Basevi’s remark is discussed in Giorgio Sanguinetti, “‘La vera analisi delle melodie’: La teoria ‘meloarmonica’ di Abramo Basevi,” in Marco Capra, ed., Una piacente estate di San Martino: Studi e ricerchi per Marcello Conati (Lucca: Libreria musicale italiana, 2000), 261–85; 275–76.
Return to text

25. Chusid, “Rigoletto and Monterone: A Study in Musical Dramaturgy,” in IMS Report 1972 (Copenhagen, 1974), 325–36; rept. in Verdi: Bollettino dell’istituto di studi verdiani 3, no. 9 (1982): 1544–58. See also Chusid’s “The Tonality of Rigoletto,” in Analyzing Opera, 241–61, and Marcello Conati, Rigoletto: Un’analisi drammatico-musicale (Venice: Masilio, 1992), 147–61 (“Aspetti della struttura musicale”).
Return to text

26. Powers, “Il ‘do del baritono’ nel ‘gioco delle parti’ verdiano,” in Bianca Maria Antolini and Wolfgang Witzenmann, eds., Opera & Libretto II (Florence: Olschki, 1993), 267–81; Rosen, “Reprise as Resolution in Verdi’s Messa da Requiem,” Theory and Practice 19 (1994): 105–120.
Return to text

27. See Joseph Straus’s discussion in Remaking the Past: Musical Modernism and the Influence of the Tonal Tradition (Cambridge: Harvard University Press, 1990), 149–55; Stravinsky’s remark is quoted on p. 152. Whether Stravinsky’s assessment of his Serenade is accurate beyond the opening measures is a question that cannot be answered here.
Return to text

28. LaRue, “Bifocal tonality: An explanation for ambiguous Baroque cadences,” in Essays on music in honor of Archibald Thompson Davison (Cambridge: Harvard University Department of Music, 1957), 173–84; rept. in Journal of Musicology 18, no. 2 (2001): 283–94.
Return to text

29. Schenker, Harmonielehre (Vienna: Universal, 1906), 352–55; Harmony, trans. Elisabeth Mann Borgese (Chicago: University of Chicago Press, 1954), 265–68. The English edition omits some pertinent annotations in Schenker’s musical examples. In line with the abstract view of Stufen taken in Harmonielehre, Schenker labels III merely as III.
Return to text

30. Compare the passage from Riemann’s Katechismus der Harmonielehre quoted in Kopp, 81–82.
Return to text

31. From a review in the Wiener allgemeine musikalische Zeitung, quoted in Pachovsky op. cit., 37. “So lässt man gewöhnlich der Stretta des ersten Finale ein sehr langsames Tempo in einem finstern, aber zugleich theatralisch wirksamen As Dur vorausgehen, worauf die lärmende Stretta im hellen C folgt.”
Return to text

32. The Overture does not figure into the larger plan of Act 1, perhaps because it was composed for Rossini’s Elisabetta, regina d’Inghilterra (1815), and was pressed into service for Il barbiere.
Return to text

33. See Ethan Haimo, “Remote keys and multi-movement unity: Haydn in the 1790s,” Musical Quarterly 74, no. 2 (1990): 242–68.
Return to text

34. The pas de trois and Tyrolean chorus (No. 15) was briefly shifted from Act 3 to Act 1 and back again during the initial performances in August 1829. This G-major number, which includes C major and E minor as secondary keys, fits well in either context. See the Fondazione Rossini’s critical edition of Guillaume Tell, ed. M. Elizabeth C. Barlet (Pesaro, 1992), 1: xxxiii–xxxvii.
Return to text

35. Weber’s chart has been the basis of recent studies of Schumann’s Dichterliebe by Fred Lerdahl (in Tonal Pitch Space) and Berthold Hoeckner (“Paths through Dichterliebe,” 19th-Century Music 30, no. 1 [2006]: 65–80).
Return to text

36. Hoeckner, 71.
Return to text

37. In the progression I–III–V, the V need not be tonicized. On the progression V–III–I see David Beach, “A recurring pattern in Mozart’s music,” Journal of Music Theory 27, no. 1 (1983): 1–29.
Return to text

38. The second-act finales of Rossini’s Ermione (1819) and Maometto II (1820) also lack tonal closure. Both operas are in two acts.
Return to text

39. Some published scores give the chord in question as G major.
Return to text

40. According to the critical edition, the upper note of the trill is G except in m. 39, where it is G. This G is missing in some editions.
Return to text

41. According to Krebs, Schubert experimented with directional tonality and tonal pairing in several songs composed in the years 1815–17, although several scholars (including Krebs himself) have offered monotonal readings of “Meeres Stille,” his sole example from 1815. While Rossini seems not to have used these techniques as early as this, these are precisely the years during which Schubert composed some of his most overtly Italianate works, such as the Sixth Symphony and the two Overtures in Italian Style. Presumably Rossini was not the only Italian composer who influenced Schubert, a student of Salieri. Hascher argues that the modulatory style of Schubert’s sonata expositions changed abruptly in 1817, and that the change involves the use of chromatic mediants as middleground harmonies.
Return to text

42. Specifically the Finale to Act 1 and the Septet in Act 2.
Return to text

43. Budden, The Operas of Verdi (New York: Oxford University Press, 1973–81), 1:16.
Return to text

44. On the substantial differences between the autograph and first edition of Il pirata, see Philip Gossett’s introduction to the facsimile of the former (New York: Garland, 1983).
Return to text

45. “Ascolta” is the final word of Gualtiero’s scena in the libretto. In Bellini’s setting, the word is delivered within the musical space of the aria.
Return to text

46. There are four ways of executing this dominant-preparation bait-and-switch; by “dominant preparation” I mean an emphatic V or V⁷ (root position) followed by a caesura. Using M for major and m for its relative minor, the four patterns are: (1) V/m to I/M; (2) V/M to I/m; (3) V/m to V/M; (4) V/M to V/m. (1) is by far the most common; it constitutes a III–I progression in the major key. (2) is relatively uncommon, probably because it resembles a conventional deceptive cadence, V–VI, in which VI is subsequently stabilized as a key; Italian composers preferred VI in such cases. (3) is a favorite of Bellini; if the second V resolves to I, as it usually does, one can hear the entire progression as III–V–I in the major key, despite the presence of caesuras after one or both dominants. (4) is the pattern heard in ex. 7; relatively uncommon, it can be heard in retrospect as VII–V–I in the minor key. A magnificent example of (4) is the link from Amelia’s scena to her aria in Act 2 of Verdi’s Un ballo in maschera (1859), in which the second dominant preparation consists of a solo for English horn, unaccompanied. See Powers, “‘La dama velata’: Act II of Un ballo in maschera,” in Martin Chusid, ed., Verdi’s Middle Period, 1849–1859 (Chicago: University of Chicago Press, 1997), 273–336; 284–85 and 298.
Return to text

47. Schenker, Free Composition, trans. and ed. Ernst Oster (New York: Longman, 1979), 26 and Fig. 13.
Return to text

48. See, however, the excellent discussion in Gilles de Van, Verdi’s Theater: Creating Drama through Music, trans. Gilda Roberts (Chicago: University of Chicago Press, 1998), 309–36 (“Unity”).
Return to text

49. The centrality of the major-third cycle F–A–D to Macbeth (both versions) was pointed out by Julian Budden in The Operas of Verdi, 1:312. Edward T. Cone claimed that Simon Boccanegra (both versions) is based on the cycle C–E–A; “On the Road to Otello: Tonality and Structure in Simon Boccanegra,” Studi verdiani 1 (1982): 72–98. Matthew Bribitzer-Stull has cited Beethoven’s Fidelio as another opera based in large part on C–E–A; see “The A–C–E Complex: The Origin and Function of Chromatic Major Third Collections in Nineteenth-Century Music,” Music Theory Spectrum 28, no. 2 (2006): 167–90.
Return to text

50. See Budden, The Operas of Verdi, 2:70; Petrobelli, “Toward an Explanation of the Dramatic Structure of Il trovatore,” Music Analysis 1, no. 2 (1983): 129–41; William Drabkin, “Characters, Key Relations, and Tonal Structure in Il trovatore,” Music Analysis 1, no. 2 (1983): 143–53.
Return to text

51. The divisions of Verdi’s operas into numbers are reflected in his autograph manuscripts but are misrepresented in many published scores. See Chusid, A catalog of Verdi’s operas (Hackensack, NJ: J. Boonin, 1974) for an authoritative listing for every Verdi opera.
Return to text

52. See Jeffrey Kurtzman, The Monteverdi Vespers of 1610: Music, context, performance (Oxford: Oxford University Press, 1999), 202–9 and 242–45.
Return to text

53. Joachim Burmeister, Musical poetics (1606), trans. Benito Rivera (New Haven: Yale University Press, 1993).
Return to text

54. Compare Carl Schachter’s discussion of similar conflicts in Chopin’s music in Unfoldings, ed. Joseph N. Straus (New York: Oxford University Press, 1999), 9.
Return to text

55. See Mary Jane Phillips-Matz, Verdi (Oxford: Oxford University Press, 1993), 167 and 701.
Return to text

56. Bailey, “An Analytical Study,” 116; Schoenberg, Theory of Harmony, trans. Roy E. Carter (Berkeley: University of California Press, 1978), 389. Compare Schenker’s expansive concept of Mischung (“combination” or “mixture”) in Harmonielehre.
Return to text

57. Daniel Heartz, “The Beginnings of the Operatic Romance: Rousseau, Sedaine, and Monsigny,” Eighteenth-Century Studies 15 (1981): 149–78.
Return to text

58. See especially Heartz’s Figures 1 and 5 (pp. 152 and 164).
Return to text

59. Rossini follows Mozart’s practice in two stage songs that belong to the tradition of the operatic romance: Almaviva’s serenade to Rosina in Il barbiere di Siviglia and Desdemona’s Willow Song in Otello (both 1816). Both dissolve into recitative without coming to a formal end.
Return to text

60. I have discussed the importance of ambitus and octave species to a nineteenth-century composition in my article “Circular Motion in Chopin’s Late B-Major Nocturne (op. 62, no. 1),” in L. Poundie Burstein and David Gagné, eds., Structure and Meaning in Tonal Music: Festschrift in Honor of Carl Schachter (Stuyvesant, NY: Pendragon Press, 2006), 19–32.
Return to text

61. Balázs Mikusi, “Mendelssohn’s ‘Scottish’ Tonality?” 19th-Century Music 29, no. 3 (2006): 240–60.
Return to text

62. See Mendelssohn’s descriptions, in his letters from Italy, of the plainchant he heard sung in the churches of Rome.
Return to text

Schenker, Harmonielehre (Vienna: Universal, 1906), 352–55; Harmony, trans. Elisabeth Mann Borgese (Chicago: University of Chicago Press, 1954), 265–68. The English edition omits some pertinent annotations in Schenker’s musical examples. In line with the abstract view of Stufen taken in Harmonielehre, Schenker labels III merely as III.

In the progression I–III–V, the V need not be tonicized. On the progression V–III–I see David Beach, “A recurring pattern in Mozart’s music,” Journal of Music Theory 27, no. 1 (1983): 1–29.

Some published scores give the chord in question as G major.

According to the critical edition, the upper note of the trill is G except in m. 39, where it is G. This G is missing in some editions.

There are four ways of executing this dominant-preparation bait-and-switch; by “dominant preparation” I mean an emphatic V or V⁷ (root position) followed by a caesura. Using M for major and m for its relative minor, the four patterns are: (1) V/m to I/M; (2) V/M to I/m; (3) V/m to V/M; (4) V/M to V/m. (1) is by far the most common; it constitutes a III–I progression in the major key. (2) is relatively uncommon, probably because it resembles a conventional deceptive cadence, V–VI, in which VI is subsequently stabilized as a key; Italian composers preferred VI in such cases. (3) is a favorite of Bellini; if the second V resolves to I, as it usually does, one can hear the entire progression as III–V–I in the major key, despite the presence of caesuras after one or both dominants. (4) is the pattern heard in ex. 7; relatively uncommon, it can be heard in retrospect as VII–V–I in the minor key. A magnificent example of (4) is the link from Amelia’s scena to her aria in Act 2 of Verdi’s Un ballo in maschera (1859), in which the second dominant preparation consists of a solo for English horn, unaccompanied. See Powers, “‘La dama velata’: Act II of Un ballo in maschera,” in Martin Chusid, ed., Verdi’s Middle Period, 1849–1859 (Chicago: University of Chicago Press, 1997), 273–336; 284–85 and 298.

The centrality of the major-third cycle F–A–D to Macbeth (both versions) was pointed out by Julian Budden in The Operas of Verdi, 1:312. Edward T. Cone claimed that Simon Boccanegra (both versions) is based on the cycle C–E–A; “On the Road to Otello: Tonality and Structure in Simon Boccanegra,” Studi verdiani 1 (1982): 72–98. Matthew Bribitzer-Stull has cited Beethoven’s Fidelio as another opera based in large part on C–E–A; see “The A–C–E Complex: The Origin and Function of Chromatic Major Third Collections in Nineteenth-Century Music,” Music Theory Spectrum 28, no. 2 (2006): 167–90.

Copyright Statement

Copyright © 2008 by the Society for Music Theory. All rights reserved.

[1] Copyrights for individual items published in Music Theory Online (MTO) are held by their authors. Items appearing in MTO may be saved and stored in electronic or paper form, and may be shared among individuals for purposes of scholarly research or discussion, but may not be republished in any form, electronic or print, without prior, written permission from the author(s), and advance notification of the editors of MTO.

[2] Any redistributed form of items published in MTO must include the following information in a form appropriate to the medium in which the items are to appear:

This item appeared in Music Theory Online in [VOLUME #, ISSUE #] on [DAY/MONTH/YEAR]. It was authored by [FULL NAME, EMAIL ADDRESS], with whose written permission it is reprinted here.

[3] Libraries may archive issues of MTO in electronic or paper form for public access so long as each issue is stored in its entirety, and no access fee is charged. Exceptions to these requirements must be approved in writing by the editors of MTO, who will act in accordance with the decisions of the Society for Music Theory.

This document and all portions thereof are protected by U.S. and international copyright laws. Material contained herein may be copied and/or distributed for research purposes only.

Prepared by Brent Yorgason, Managing Editor and Cara Stroud and Tahirih Motazedian, Editorial Assistants
Number of visits: