John Michael Cooper

One Aria or two? Mendelssohn, Metastasio, and Infelice

In 1851, as part of the first phase of posthumous publications of works by Felix Mendelssohn Bartholdy, Julius Rietz edited and published with the Leipzig firm of Breitkopf & Härtel a “Concert Arie für die Sopran–Stimme mit Begleitung des Orchesters,” assigning the aria the posthumous opus number 94. Beginning in 1853, inventories of the composer’s works added the title “Infelice,” extracting this title from the text incipit of the recitative portion of the work.[1] This information was repeated in the catalog of Mendelssohn’s published works released by Breitkopf & Härtel in 1873, but the press now added that the aria had been composed in 1834 and revised in 1843, also identifying Metastasio as the poet.[2] In 1876 Rietz reissued his edition in Series 15 of the edition of Mendelssohn’s collected works published by Breitkopf & Härtel.

Thus, since the 1870s all editions and recordings of this aria, as well as catalogs of Mendelssohn’s output, have consistently identified this music by the recitative’s text incipit (“Infelice”) and ascribed its initial composition to the year 1834, at the same presenting the work as it existed in 1843. The Op. 94 aria has been viewed as the later of two manifestations of a single compositional concept and thus vested with greater compositional authority, in accordance with the premises of the editorial philosophy of the Fassung letzter Hand. In so doing, these posthumous commentaries, editions, and performances have placed this aria in the company of a sizeable number of works by Mendelssohn whose genesis spanned many years, and whose latest versions the composer explicitly granted artistic primacy as the “best” versions of works whose earlier renderings had but imperfectly realized his compositional aims.

At first blush, the traditional understanding of Mendelssohn’s Op. 94 has much to commend it. Both incarnations of this compositional concept set poetry by Metastasio; both are written for female voice with orchestra; both are in B–flat major; both employ the well–known form of “scena and aria”; and both include a brief but prominent return of the music from the cavatina near the end of the cabaletta (after the model of Beethoven’s Ah, Perfido!). Finally, the music of the recitative of the 1843 composition is clearly derived from that of its 1834 counterpart, and the essential thematic material of the cavatina is almost identical.

Despite these similarities, however, there is no evidence that Mendelssohn and his contemporaries considered the 1843 aria as a revision of the 1834 one. Rather, contemporary reviews and the composer himself referred to the 1843 composition as “new.”[3] Moreover, there are major extensive differences in the two works’ texts, circumstantial origins, music, and large–scale interpretive gestures. These differences suggest that while Mendelssohn would have recognized both the 1834 aria and its 1843 counterpart as abbandonata–style concert arias for female voice with orchestra, he surely would have considered them as autonomous compositional responses to their respective texts.

The present remarks trace the two arias’ histories, identify several salient similarities and differences, and comment on some significant implications of the divergences between the two. Ultimately, I propose that Mendelssohn’s musical maturity produced not one concert aria based on texts by Metastasio, but two – and that this realization offers to the musical world an entirely new composition whose identity has for more than 150 years been wrongfully subsumed into that of another work.[4]

Composition

The first of Mendelssohn’s mature Metastasio settings resulted from an 1832 commission from the Philharmonic Society of London for an overture, a symphony, and a vocal piece.[5] Although Mendelssohn completed two of the commissioned works in the spring of 1833, he was unable to begin work on the vocal piece until February 1834. By that point, he had decided that the work would be a setting of verses by Metastasio, and would be clearly tied to two of the leading figures of contemporary musical life. The composer wrote to his family on February 19, 1834:

… my vocal scena for the Philharmonic will be finished in a few days. The text is the most beautiful nonsense by Metastasio (recitative, adagio, and allegro) assembled from four different operas – but all that should be made good again by a solo violin which accompanies the voice, and for which I’m speculating on de Bériot.[6]

As may be obvious, “de Bériot” is a reference to the contemporary violin virtuoso Charles–Auguste de Bériot (1802-70), who at the time was engaged in an extramarital affair with renowned soprano Maria Malibran. The importance of this relationship is indicated in Fanny Hensel’s response to her brother’s letter a few days later:

Concerning your vocal scena you write only that it contains an obbligato violin part for Beriot [sic]; from that we surmise that you’re thinking of Malibran as the soprano. Does the key fit?[7]

Mendelssohn completed the aria on 3 April 1834. On 14 May his London friend Karl Klingemann reported that Maria Caradori-Allan (1800-65) would be the soprano soloist, and the premiere occurred at the Philharmonic Society’s concert on May 19, with concertmaster J. D. Loder as violin soloist and Thomas Cooke conducting.[8] A second performance was given on 11 April 1836 with the same orchestra and soloists under the direction of Ignaz Moscheles. This performance, given just two weeks after the marriage of Malibran and de Bériot, was the last in the composer’s lifetime.

Thus, Malibran and de Bériot never performed the 1834 aria despite the important role they played in its conception. Although it most likely was in good hands in its performances by Caradori-Allan, she possessed neither the brilliant technique nor the popular fame of the intended soloists on whom Mendelssohn had been “speculating” when he wrote the piece. Moreover, the well–known romance between the two soloists would have granted the aria an undeniable topical appeal. This circumstantial influence is significant, for as will be shown presently, Mendelssohn’s setting of Metastasio’s text develops an elaborate and obviously amorous musical relationship between the two soloists which would have paralleled their real–life relationship.

As shown in Table 1, there are three surviving manuscripts for Mendelssohn’s 1834 Metastasio aria. Mendelssohn had the first of these – the autograph composing score, dated at the end “Düsseldorf den 3ten April 1834” – bound ca. 1839 into what would become Volume 28 of the Mendelssohn Nachlaß.[9] As shown in Figure 1, this autograph reveals extensive revisions to both the text and the music of the aria. A second manuscript, produced by Mendelssohn’s favorite copyist in Düsseldorf, J. G. Schauseil, is found in the library of the Royal Academy of Music in London; this manuscript also reveals several revisions in Mendelssohn’s own script, primarily in matters of text setting.[10] Finally, the Royal Academy of Music library also contains a reduction for voice and piano. This manuscript is in the script of William Goodwin, a copyist who had worked with the Philharmonic Society since at least 1829.[11]

By contrast, documentation pertaining to the 1843 aria is sparse. It exists in only one known contemporary manuscript: the autograph full score that would be bound in volume 38[12] in the Mendelssohn Nachlaß found in the Biblioteca Jagiellońska, Kraków, sometime after November 1844; its final page is dated 15 January 1843. The work was performed at an “Extraconcert” given by Sophie Schloß in the Leipzig Gewandhaus on 9 February 1843, and two more performances were given during Mendelssohn’s lifetime; both featured Schloß as soloist. This aria was posthumously published in 1851 by Breitkopf & Härtel in Leipzig (with Italian and German text) and Ewer & Co. in London (with Italian and English text).

Mendelssohn mentioned the 1843 aria in two surviving letters. The first of these was written to his brother Paul on 17 January 1843, just two days after the completion of the score:

I want to tell you about the programs of our upcoming concerts; perhaps one or another of them will appeal to you or Albertine [Paul’s wife]… On 9 February there is the concert by Mlle [Sophie] Schloß in which Mme [Clara] Schumann and [Gewandhaus concertmaster Ferdinand] David will also perform, individually and together, and in which a new concert aria by me which I have written for Mlle Schloß will be given, and probably also my Meeresstille…[13]

Mendelssohn again referred to the work in a letter to his friend Ferdinand Hiller written on 3 March, and here he makes explicit that the 1834 work was the historical starting point for the same aria that he had described as “new” in his letter of 19 January:

I sought out the Scena for Mlle Schloß’s benefit concert, wrote a new Allegro for it, and thereby did my part. Beyond that it is of little use.[14]

Since contemporary reviews published in the Leipzig Allgemeine musikalische Zeitung and the Neue Zeitschrift für Musik clearly concur with the composer in describing the 1843 aria as “new,”[15] and since these reviewers were clearly responding to information provided in the program itself, we may safely conclude that the work was billed as “new” at the time. This evidence contradicts the conventional assumption that the 1843 aria represented a revision of its 1834 counterpart, – a reworking that was consequently more authorized. Indeed, by stating that the 1843 aria was “of little use” Mendelssohn denied that aria’s artistic merit and critical authority.

Text

An examination of the texts and textual sources of these two arias immediately raises problems for the conventional understanding of the relationship between them, since compositions with any amount of substantively different similar texts are generally considered either as contrafacta or as revisions of the same textual/musical concept. In order to be designated contrafacta compositions must employ essentially the same music for all textually divergent portions; and in order to be considered textual/musical revisions the text must be essentially the same in the two versions. Neither of these criteria fits well in the instance of these two arias: the music for most of the arias (recitative, cavatina, and cabaletta alike) differs far more than is permissible under the rubric of “contrafacta” and the textual divergences are too great and too substantive to permit classification as a “revision” according to normal standards. In other words, from a textual perspective Mendelssohn’s 1843 Infelice cannot be considered the same work as the 1834 aria.

As shown in Table 2, Mendelssohn compiled the texts for the two arias from several different opera seria libretti by Metastasio. The original source of the titular incipit is unknown; it may well stem from Mendelssohn himself, or from any number of other Italian operas. After that exclamation, the recitative of the 1834 aria presents excerpts from Il Trionfo di Clelia, III/3; Romeolo ed Ersilia, III/5; and Giustino IV/7, with the interpolated line “e l’amo pur” from some other textual source that remains unidentified; these texts are reused, with slight modifications and the additional interpolation “E pur odiar nol posso ancor!,” in the 1843 aria. The cavatina of the 1834 aria returns to Il Trionfo di Clelia and Giustino, while the 1843 aria retains only lines 15, 20, and 21, replacing the other lines with more material from an unidentified source. Finally, the cabaletta of the 1834 aria returns to Giustino for the entirety of its text, while the cabaletta of the 1843 aria retains only three of the lines from that opera, replacing the remainder with material from an unidentified source. It is possible, of course, that these untraced lines come from more than one source, or that they were penned by Mendelssohn himself.

These textual interrelationships make clear that Mendelssohn’s 1834 Metastasio aria offered no more than a compositional Auftakt for the text of the later one – for in selecting and/or writing the various replacement texts that are present in the later work, the composer probably had to return to his source of Metastasio’s poetry and find new text, with the full knowledge that this new text would also lead to new music. This approach normally dictates that the music written for the new text grant to the composition the status of an autonomous work.

Music

An examination of the music of the two arias corroborates this impression. As shown in Examples 1a and 1b, the recitatives of two arias employ similar harmonic plans, but the later aria dispenses with the fugato textures of the earlier one, replacing the agitation suggested by the doubled cascading eighth notes in the 1834 aria instead with a series of throbbing syncopated harmonies in the later strings.

An instance of more pronouncedly divergent parallelisms obtains in the two cavatinas: As shown in Examples 2a–b, the two employ the same essential thematic material. This material comprises two ideas: the signature gesture of the first comprises a triadic descent leading to the dominant of the subdominant, while that of the second comprises a set of dotted–eighth and sixteenth–note embellishments of the dominant of the new key (F major).

The most prominent musical divergences, however, occur in the cabaletta. As shown in Example 3a and 3b, the 1834 Allegro is based on five discrete melodic ideas, and these bear little or no affinity to the three ideas on which the 1843 Allegro is based. The disparity is, of course, only natural, for as noted above the texts of the cabaletta sections are completely different.

Interpretation

Finally, the music of the two arias reflects Mendelssohn’s substantially different compositional interpretations of Metastasio’s poetry. As already noted, the music and text of the recitatives differ but little: both introduce the persona of the poem’s love-forsaken protagonist and set her words to music that bespeaks her psychological and emotional turbulence. In the cavatina, however, the works unfold in a substantially different fashion. In the 1834 aria Mendelssohn also introduces the persona of the lyric persona’s lost love, allegorized in a finely wrought obbligato violin part (which, as noted above, would have been personified by de Bériot). Indeed, it is this lover whom we first encounter, for the cavatina begins with a presentation of its lovely Italianate main melody in the solo violin (Ex. 4a). After a full cadence, this melody is taken over by the protagonist, wistfully remembering the “età dell’oro” when her lover was still present (Ex. 4b); she then loses herself in the memory of those happy days, when “a tender shrub and a limpid stream nourished the people.” Finally, the two personae are united in an elaborate and finely wrought duet, the solo violin alternately weaving graceful countermelodies around the soprano lines and, toward the end, moving in parallel thirds and sixths with those lines (Ex. 4c).

The elaborate development involved in this meticulously crafted Andante deserves comment for two reasons. First, in creating two different personae – one associated with the solo soprano, the other with the solo violin – Mendelssohn identified the work with its intended performers; for the extramarital affair of Malibran and de Bériot was well known by 1834, and a performance of the work featuring the two of them in such conspicuously amatory solo roles inevitably would have invited contemporary audiences to identify the textual and musical personae with the performers themselves. Second, since the development of the lovers’ relationship graphically depicted in the Andante is entirely absent (or at best an unstated thing of the past) in Metastasio’s text, the creation and development of that relationship was obviously an important consideration in Mendelssohn’s composition of the aria. Such gestures are by nature assertions of a composer’s identity, confirmations of a Romantically individualized reading of the original text.

Now, there is no doubt that the Cavatina of the 1843 aria takes its 1834 counterpart as its starting point, since it reuses the essential thematic material (Ex. 5). Nevertheless, the 1843 cavatina is markedly shorter and includes no violin obbligato. Consequently, this 1843 Andante does not graphically develop the lovers’ past relationship, but remains firmly grounded in the present; indeed, rather than becoming lost in her memories this speaker is consciously aware that she is reminiscing, especially in lines 19–22: “Ah, when I look inside myself I am always reminded of the day he vowed to be true to me.”

Likewise significant are the widely differing compositional interventions Mendelssohn made in the cabalettas of the two arias. Initially, the 1834 cabaletta returns to the comparatively straightforward style of text setting established in the recitative, and the Allegro as a whole is characterized by tonal instability and angular melodic lines – musical gestures conventionally associated with the affective turbulence of the soprano persona’s thoughts. As shown in Example 6, however, Mendelssohn introduces into the 1834 cabaletta a conspicuously static discursion into D minor. This D-minor interlude constitutes a reminiscence episode in which the solo violin and the soprano, reunited in dialog, resume the pastoral text and texture of the Andante wherein they were first joined: “the world was happy then, when a tender plant [and] a limpid brook nourished the people” (ll. 18-21). The reminiscence implicitly becomes more real as the music returns to F major, the key from which it initially departed, but the tonality abruptly turns to F minor in measure 258, and the return of the agitato running doubled eighth notes in the strings at that point emphasizes that the episode was only a fleeting daydream, far removed for the bitterness of the present.

The end of the reminiscence episode seems to mark a crucial recognition for the speaker in the 1834 aria and a dramatic turning point in the composition, for the agitato mood prevails to the end. The text with which the Allegro began (“D’amore nel regno” etc.) is now set as a series of short, breathless, modulatory phrases rather than a bona fide melody, and the restatements of other material are all modified to heighten tension, primarily through increased harmonic instability. This section finally dissolves into a crescendo in which the solo soprano repeatedly exclaims “disrupt it again” (“la turba ancor”) in appropriately fragmented phrases. As shown in Example 7, the climax of this intensification – a tonic 6/4 chord with a fermata – is followed by a final brief return of the music and text of the Andante featuring both the solo soprano and the solo violin – a reflective duet–entreaty for a return of that “bell’età.” The plea remains unanswered, however, and the aria closes with a series of reiterated cries of “Ah, ritorna!,” followed by a final turbulent statement of the doubled–eighth–note figure.

In short, although Mendelssohn’s 1834 aria responds sensitively to the salient affects of Metastasio’s texts, it also represents both his own highly personalized reading of that text and a setting that would have possessed considerable topical appeal for a musical public whose familiarity with Metastasian aesthetics was at best limited.

The 1843 aria Infelice! / Ah, ritorna, età felice likewise reveals much about Mendelssohn’s compositional agenda and the context for which it was written, but its treatment of Metastasio’s text differs considerably – not least of all, of course, because the two cabaletti employ different texts. As already noted, the thematic material is unrelated to that of the 1834 aria. Moreover, the 1843 aria offers no reminiscence episode in the cabaletta. Most important, however, is that the 1843 aria offers a final resolution entirely absent in the 1834 composition: “And yet just the memory of those love–filled days can soothe this bitter grief!” (ll. 29-32). Mendelssohn evidently recognized this exclamation’s potential for resolving emotional conflict, for in the first part of the 1843 cabaletta he cultivates an unstable, dramatically charged musical style that emphasizes the protagonist’s emotional tension (“B” in Example 3b, above), but in the final bars he musically suggests that the consolation possible through memory has in fact been attained: the second theme from the main Allegro, formerly in an unstable G minor, is now presented firmly in B-flat major, as the subject of the coda (Ex. 8). The consolation so emphatically denied in the earlier work is thus granted in the 1843 aria – a Romantic interpretive intervention that transforms the essentially static composite text into an end–weighted gesture whose goal and culmination is the resolution of the speaker’s emotional conflicts in the final measures.

There is, in fact, much more to be said about these two largely unknown arias. Most important for purposes of this paper, however, is the insight the works provide into changes in Mendelssohn’s view of the public composer’s role vis-à-vis that of the poet in dramatic works. For while both arias present Romantic compositional reinterpretations of Metastasio’s texts, the two works also represent conspicuously different points in his career. The 1834 aria, written during his first professional engagement and at the beginning of what promised (but was by no means certain) to be an illustrious public career as a musician, deliberately cultivates association with established celebrity: Beethoven’s Ah, perfido!, the glamorous Malibran and de Bériot, and a stylistic idiom overtly indebted to the lyrical drama of contemporary Italian opera. By contrast, when he composed the 1843 aria Mendelssohn stood at the early height of his international prestige as public figure, composer, and conductor. The aria acknowledges this different relationship between poet, composer, and musical public not only through the more intrinsically dramatic mezzo-soprano tessitura of its vocal soloist, but also by its emphasis on a last-minute climactic resolution of dramatic conflict. The two works thus represent contrasting approaches to the quintessential Romantic issue of the role and artistic obligations of the composer as interpreter and dramatizer of poetry – and Metastasio, however unwittingly, played a central role in Mendelssohn’s struggle with the nature of Romantic artistic identity.

In conclusion, we must ask whether these two arias represent an instance of compositional revision, or simply of the recycling of musical material between two generically similar works. If the former is the case, the implications are that Mendelssohn reused material from the earlier composition in a later one; that he presumably considered the later work as an improved realization of the concept first executed in the earlier aria; that he viewed the later work not as “new,” but as “old”; and that he intended the later work to supersede the earlier one. In this instance, the two Infelice arias would emerge as counterparts to Mendelssohn’s two settings of Goethe’s Walpurgisnacht. In this view, however, we must concede that the Infelice arias occupy a troublingly exceptional status: for the textual and musical differences between the two arias normally would automatically grant them musical autonomy according to conventions of musical identification, and the musical differences between these two versions are significantly greater than those in any other known instance of Mendelssohn’s extensive revisions. Finally, we must remember that the 1843 concert aria, unlike the Walpurgisnacht, remained unpublished until after the composer’s death: Mendelssohn never granted the later version greater authority than the earlier one.

On the other hand, if the 1843 aria represents a recycling of some text and music in the context of an ultimately autonomous composition, this has significant consequences. This view is appealing not least of all for methodological reasons, since it accounts for the textual and musical variances between these two settings in a fashion that is consistent with accepted practices for identifying texted musical works (consider, for example, the conventions of identifying and separating out works using “Ave Maria” texts). Moreover, such an approach places Mendelssohn’s two Infelice arias in a group not with the Walpurgisnacht and the “Scottish” Symphony, but with his two settings of “Was betrübst Du dich, meine Seele, und bist so unruhig in mir? Harre auf Gott; denn ich werde ihm noch danken, daß er meines Angesichts Hilfe und mein Gott ist.“ This verse, stated in both Psalm 42 and Psalm 43 in the Lutheran Bible, naturally figures in Mendelssohn’s settings of those two works. His setting of Psalm 42 was composed in 1837 and published Op. 42 in 1838/39, and his setting of Psalm 43 was composed in 1844 and posthumously published as Op. 78 No. 2 in 1849. Although they are scored for radically different ensembles, both settings not only share this text, but also use the same music for all shared material – yet unlike the two Infelice arias, the psalm settings are conventionally assigned discrete identities in inventories of Mendelssohn’s oeuvre. Finally, this view liberates the 1834 aria from the anachronistic shadow of its 1843 counterpart, shedding new light on an otherwise entirely unknown composition that represents Mendelssohn’s first masterful essay in the early nineteenth–century Italian operatic style.