Verdi Forum

Number 12 Article 4

1-1-1984 Text Setting in Verdi's and Jeffrey Langford Manhattan School of Music

Follow this and additional works at: http://scholarship.richmond.edu/vf Part of the Musicology Commons

Recommended Citation Langford, Jeffrey (1984) "Text Setting in Verdi's Jérusalem and Don Carlos," Verdi Forum: No. 12, Article 4. Available at: http://scholarship.richmond.edu/vf/vol1/iss12/4

This Article is brought to you for free and open access by UR Scholarship Repository. It has been accepted for inclusion in Verdi Forum by an authorized administrator of UR Scholarship Repository. For more information, please contact [email protected]. Text Setting in Verdi's Jérusalem and Don Carlos

Keywords , Jérusalem, Don Carlos

This article is available in Verdi Forum: http://scholarship.richmond.edu/vf/vol1/iss12/4 Text Setting in Verdi's Jerusalem and Don Carlos

Jeffrey Langford, Manhattan School of Music

In the summer of 1847 Verdi went to Lon­ poor French was dealt with by the fact that don to supervise the premiere of I Masna­ while writing the he would be living dieri at Her Majesty's Theatre. This was in Paris where he could easily get help his first premiere outside of , and his from friends like Strepponi or from his li­ second international trip of any kind . On brettists. 3 Another impediment, his busy returning home from London, Verdi schedule, was removed through an agree­ stopped in Paris to visit his French pub­ ment with the Opera allowing the com­ lisher, Leon Escudier, and Giuseppina poser to reset an earlier to a Strepponi, who had been teaching voice new French . I Lombardi was se­ there since the fall of 1846. During this lected for this transformation, and the job visit Verdi was approached by the Paris Op­ of adapting the drama and drafting the new era and asked to write a new work on a libretto was entrusted to Gustav Vaez and French text. Althouyh such requests had Alphonse Royer.4 Their work resulted in come to him before, he had always been much more than a simple translation of the forced to refuse them, for a variety of rea­ Italian libretto into French. While carefully sons. For one, he was busy fulfilling Ital­ designed to allow Verdi to reuse most of his ian commissions and had no time for for­ earlier music, the new libretto, entitled Je­ eign theaters. In addition, he was not fluent rusalem, actually involved major changes in French and undoubtedly had reserva­ in I Lombardi. s tions about composing a major work for Verdi's work on Jerusalem represented a Europe's leading in a language venture into compositionally uncharted with which he was not completely secure. 2 waters. In Italy his exposure to French op­ In these circumstances, one wonders era could only have been minimal. The why, in the summer of 1847, he suddenly chronicles of such opera houses as La agreed to accept the Opera commission. Scala, , and the Regio in Parma He was, after all, still busy with other Ital­ show that most French never ian commissions, and his French was no reached the theaters of major Italian cities; more fluent than before. The answer may and those few works that did (most notably involve a combination of factors . Verdi's Meyerbeer's Robert le Diable and Auber's

'See his letter of August 1846 to Vincenzio documentary evidence in the form of letters to Flauto in which he says, " If none other, the indicate that he ever actually approached them of Paris would not disdain to open for help. The presence of Strepponi in Paris al its doors to me , as I can show you with a letter that time may well have been the decisive factor from Pillet. " Gaetano Cesari and Alessandro in his decision to write Jerusalem (eds.). Luzio, eds., I copialettere di Giuseppe Verdi 'Three of Verdi 's most successful early operas (, 1913) 24 . (those predating 1846) had already been staged 2As of 1847 Verdi spoke French well enough to in Paris: , 16 Oct. 1845; Emani, 6 Jan. get by in Paris, but not until the 1860s did he 1846; I due , 17 Dec. 1846. Only I Lom­ feel comfortable enough to write an occasional bardi had not yet been mounted there. A photo­ letter to Escudier or Du Locle in French. Even copy of Verdi 's contract for Jerusalem with the then most of his letters to French-speaking Opera may be found in the archive of the Ameri­ friends were still written in Italian. This practice can Institute for Verdi Studies at New York Uni­ contrasts markedly with that of Giuseppina versity . Strepponi, who, after she began living with sA lengthy discussion of the transformation of Verdi , always wrote to the composer's French I Lombardi into Jerusalem may be found in colleagues in their own language. David Kimbell, "Verdi's First Rifacimento: 'I 3Unfortunately, the fact that Verdi worked di­ Lombardi' and 'Jerusalem,' " Music & Letters, rectly with his librettists means that there is no VQI. 60, no . 1 (January 1979) pp. 1-36.

19 la Muette de Portie!) were always given in line of verse. 8 Italian words show consider­ Italian translation. 6 So while Verdi may able variety in the location of their natural have known some of the music of major stresses. The usual rule is that multi­ French composers, he probably remained syllable words receive a stress on the pen­ unfamiliar with the exact correlation of that ultimate syllable, as in the word amore. music with its original French texts. The This is called the normal or piano ending. composition of Jerusalem raises the inter­ In many other words, however, the stress esting question of how Verdi dealt with the falls on the antipenultimate syllable, as in actual mechanics of French text setting the word virgine. This type of word is said early in his career, and gives us the oppor­ to have a sdrucciolo ending. Lastly, many tunity to discover what he thought about words in the Italian language have a stress the effect of language on musical style. Ul­ that falls on the last syllable, as in the word timately an investigation of the subject may belta or in words with a single syllable. lead to a better understanding of the exact These words have a tronco ending. relationship between Verdi and the French In all Italian poetry the final stress of a grand opera tradition. line (the penultimate syllable in a normal, piano ending) is called the primary accent, the heaviest stress, of that line. Other stresses are referred to as secondary accents. The difference between these two The primary concern of any composer levels of accent is shown in scansion as fol­ attempting to set a versified text to music lows: = marks the primary accent, - lies in the establishment of a system for the marks a secondary accent ( ~ marks an correlation of the stresses of poetic prosody elision). with the natural accents of musical meters. Stress within complete lines of Italian In order to understand how a composer like poetry is governed primarily by poetic me­ Verdi approaches this problem we must ters (pre-existing patterns of stressed and know something about the basic principles unstressed syllables) which, in turn, are of Italian and French versification. 7 determined by the number of syllables in a In Italian poetry, individual words have line. In opera some of the most stressed and unstressed syllables which, common line lengths are quatemario (4 when taken in conjunction with the rules syllables), quinario (5), senario (6), set­ governing poetic meters, help to determine tenario (7), ottonario (8), decasillabo (IO), the placement of stress within a complete and endecasillabo ( 11) . Lines with an even

6See Pompeo Cambiasi, ed., Rappresentazioni 1973); Leon E. Kastner, A History of French date nei reali teatri di Milano, 1778-1872 (Mi­ Versification (Oxford, 1903); and Jean Suber­ lan, 1872); Paolo-Emilio Ferrari, ed., Spettacoli ville, Histoire et theorie de la versification fra­ dramatico-musicali e coreografici in Parma nfaise (Paris, n.d.). I would also like to thank dal/ 'anno 1628 al/ 'anno 1883 (Parma, 1884); professor Nicholas Granito of the Manhattan and Luigi Lianovosani (pseud. for Giovanni Sal­ School of Music for his generous advice on mat­ violi) ed., La Fenice: gran teatro di Venezia; ters concerning Italian and French prosody. serie degli spettacoli dalla primavera I '792 a 'The subject of stress in language is itself a very tutto ii camovale 1876 (Milan, 1878). complex one. Generally, stress results from the 7The following summary of the rules of Italian coordinated effect of three factors: the lengthen­ versification is digested from Luciana Castel­ ing of a vowel (agogic accent) , an increase in naovo, La metrica italiana (Milan: Vita e pen­ dynamic level (dynamic accent), and a rise of siero, 1979); Pier Enea Guarnerio, Manuale di pitch in the voice (pitch accent). The relative im­ versificazione italiana (Milan, 1913); and Rob­ portance of these elements in determining stress ert Moreen, "Integration of Text Forms and varies considerably from language to language. Musical Forms in Verdi's Early Operas," In classical Italian, stress is primarily the result (Ph.D. diss., Princeton University, 1975). The of durational accent, while in French it is pitch rules of French versification are taken from accent that contributes most to the perception of Pierre Guiraud, La Versification , 2nd ed. (Paris, stress.

20 number of syllables tend to have more reg­ I~ cietCi,_ed ~rg~rn~i ular metric patterns than those with an un­ seceuu even syllable count. For example, an ot­ Dov~ m'Ortal n~n ~- tonario may have the pattern -· - · - · - ­ (stresses on syllables 1, 3, 5 & 7) or here three different metric patterns, but we -· - -·· -- (stresses on 1, 4, & 7) or also see a combination of sdrucciolo and tronco lines adding to the effect of variety ··-·· • -- (only two stresses, on 3 and 9 7) . Generally, once a strophe of lines with and unpredictability. an even syllable count begins in one pat­ As far as prosodic flexibility is con­ tern, however, it does not change to an al­ cerned, Italian poetry usually differs mark­ ternate pattern. Example 1 illustrates part edly from French. Primarily the difference of a typical ottonario strophe from Verdi's involves the location and relative intensity I Lombardi. of secondary accents. Example 1. Pagano's cabaletta from Stress in the French language results No. 2 ,s. Act I, sc . 5, I mainly from a pitch inflection that falls on Lombardi the last spoken syllable of the final word in a line or phrase. (Mute endings of -e, -es, 6 sJ)eranzva di vend~tta and -ent are not counted as final syllables.) Individual French words are usually said to Gia sfaVliti sul ntio volto have a stress on their final syllable, but in Da ~t anni.Jl me dil~tta reality individual words have no percepti­ ble stress when considered apart from a Alt~ voce non ~s ~lw complete line or phrase to which they might belong. Within such word groups in French poetry, as in the spoken language, Verses of uneven syllable count, on the the primary accent falls on the final sylla­ other hand , show greater flexibility. A set­ ble of a line. But in French poetry, lines of tenario line, for example, may have more than eight syllables subdivide into accents that fall on syllables 2 and 6; 3 and two phrases (hemistiches) with a cesura 6; 2, 4 and 6; 1, 4, and 6; or 1, 3, and 6; and another primary accent at the end of and poets may freely mix these different 10 the first phase. Beyond the fixed location meters within any single strophe. Thus, the of these primary accents little else can be variety of stress patterns is significantly said with certainty about stress in French greater in strophes of uneven syllable poetry. The location of secondary accents count, as example 2 from I Lombardi dem­ is not determined by various pre-existing onstrates. Not only do we see metrical patterns as in Italian poetry. Such concepts of versification are completely Example 2. ll . 5-8 of Oronte's cavatina foreign to the French. Instead, the location "La mia letizia infondere," of these accents is determined by internal No. 6, Act II , sc. 2 syntactical relationships. That is to say, T~ntvsmonie n~ll ' cte~ each hemistich is capable of further subdi­ vision by lesser cesuras that mark natural Quanti pianetLegli_ha breaks in the grammatical structure of the s.Numbers are those given by Verdi himself (i .e . lines ending with tronco or sdrucciolo words are in his autograph manuscripts) and reported by counted as though they had normal piano end­ Martin Chusid in A Catalog of Verdi 's Operas ings. Thus, lines of 6, 7, and 8 actual syllables (Hackensack, N.J., 1974). may , depending upon the last word of the line, 9In Italian poetic theory the syllable count of any all be called settenario. line is predicated upon a normal piano ending. 101n the nineteenth century Victor Hugo popular­ When the weak final syllable of the piano ending ized a division of the alexandrine ( 12-syllable is missing (tronco) or when an extra weak sylla­ line) into three main sections, each with its own ble appears at the end of the line (sdrucciolo) the primary accent. See Kastner, French Versifica­ syllable count is unaffected. In other words, tion, pp. 91-94.

21 line. Each lesser cesura in turn creates a imperceptible." 14 This means that French secondary accent of its own, and both the poetry should generally be recited as a se­ number and location of these accents ries of unstressed syllables leading to a pri­ change from line to line. As Kastner ob­ mary accent at the end of each phrase or serves, "it is to this freedom in the disposi­ line. In music, of course, this theoretical tion of the accents, other than that at the ideal of only one accent per phrase cannot [medial] cesura and at the end of the line, be realized because some of those remain­ that French versification owes one of its ing unaccented syllables will inevitably fall chief advantages-the multiplicity of possi­ on strong beats of the musical meter and ble rhythmical periods and combina­ will thereby assume a stress that they might tions." 11 Example 3 illustrates this varia­ not otherwise deserve. A composer, then, tion in the placement of secondary accents is forced by the very nature of the medium in a few alexandrines from Racine's Atha- in which he works to identify the second­ • 12 [ 1e. ary accents in his libretto (weak though they may be) and to somehow coordinate Example 3. them with the strong and weak beats of his Je viens I sel~n l~u~gt:Jlj'nilque._Let musical meter. solenn~I Although stress in French poetry gravi­ tates naturally toward the irregular and Cel~brer I avec v;us I I I~ fameu+se v =v flexible patterns seen in example 3 above, journee it is important to note that French poetry Tc ~= J' ai aabi ma noble mait=ae:

22 Five days later he wrote to Du Locle again, irregularities of Verdi's French text setting this time asking for an additional strophe: are explained without resorting to such lev­ els of linguistic subtlety. In order not to waste time, I have written this strophe in Italian; if convenient, translate it into French for me, conserv­ II ing, of course, the accent and the rhythm. Requests from Verdi to his librettists for Elisabetta French prosody modeled after Italian pat­ terns suggest the process by which I Lom­ Pace, perdon dal ciel~ bardi was to be transformed into Jerusa­ I~vochefb ~r l~i. lem. The inherent variety of stress patterns in French poetry, along with the skill with V-v -v =v Pietoso~i stenda__un velo which Royer and Vaez matched the line lengths of their new French text to the orig­ S~i falfi dl cost~i inal Italian, allowed Verdi to adapt his ear­ EboIi lier music to the new libretto. An example Avh no...; __il perdon d~I cielo of such adaptation can be found in the cor­ responding arias: Giselda's "Salve Maria" Chiamar non os~~i. (/Lombardi, No. 3, Act I, sc. 6) and He­ Stender non puote_un ~I~, lene's "Ave Maria" (Jerusalem, No. 2, Act I, sc. 2). The Italian poetry consists of v - v - v = lfl Son troppi_i falli miei. eight lines of quinario doppio in which the primary accent falls on the fourth syllable Further complicating this subject of pro­ of every quinario group and a secondary sodic theory is the fact that normal word accent falls variably on either the first or stress in many languages, but especially in the second syllable. Example 5 illustrates French, is capable of great elasticity. A the first three doppio lines. principle that the French call accent ora­ toire allows normal stress patterns to be Example 5. modified by the emotion or character of the person who is speaking. As Romain Rol­ salve M'ari~-d'i graz~1 ~tto land explained to Richard Strauss, for ex­ T'empie_il si ;is';. ample, the word cheveux is normally Sig~~-che_in te stressed on the last syllable, but "when Tuo ctivin f~tto-sla bened~tto . spoken by a man in love" the word may come out CHEveux. 17 There are, according Notice that the secondary accent some­ to Rolland, "a certain number of words times shifts in the second quinario group of which have an absolutely fixed stress: they each line. Verdi's setting of this text ap­ form the skeleton of the language. The oth­ pears in example 6, from which we can in­ ers are fluid and Protean; they obey and fer several characteristic procedures of his yield to corcumstances which are logical, setting of Italian texts: 19 psychological, etc." 18 For the purposes of 1. Verdi consistently places the primary this study, however, the complexities of accents of the poetry on the strongest accent oratoire may be eliminated from beat of the measure. consideration because, as we shall see, the 2. The placement of secondary accents is

16/bid. p. 164. Scansion marks are mine. There s~ondence (Berkeley, 1968) 46. is no sign of these strophes in either the printed I Ibid. p. 47 . libretto of 1867 or in the score. The 7-syllable 19 As always, general procedures have their ex­ meter that Verdi wanted Du Lode to maintain is ceptions. Nevertheless, these characteristics are an unusual one in French poetry. consistent throughout the vast majority of Ver­ 11Richard Strauss and Romain Rolland: Corre- di's Italian text settings.

23 Example 6.

1 So.I - ve Mo.. -: ri - Q Ji _9ra-zia. ii: pd - to h111-pieil

A ~e. Mo.. - ri - o.. 11111. '/oix h. pri - e 0 Vicr-~e.

nor c.he. i~ te. ~i po - : So.. :tlJ.-0 ;\;_,.;~: r11.t-ro ~I

less predictable. In four of the five in­ servations. Giselda's cabaletta "No giusta stances in this excerpt they fall on either causa" (No. 9, Act III, sc. 9) is also in of the strong beats of the measure ( 1 quinario doppio meter; but the prosody of and 3), but in measure 5 the secondary these lines is much more regular than in the accent falls on the weak second beat previous example. while the unaccented syllable oft 'empie Example 7. occupies the strong beat-an apparent N~ ... gwsta c~u~-non ~ dlddto reversal of the normal prosodic values of the text. L~ terra s~rgere-di sangu~umano; 3. Unaccented syllables usually fall in senso weaker metric positions than do any of E. turpe_insani~-non Pio, the accented syllables that surround Chull'oro d~sUiS\-d~l mons~l~no . them. (See m. 8 for a common excep­ tion to this practice.) Queste de! c~lo-non fur pa;;;le . . . Verdi's occasional disregard for poetic No, olo no! v~ol~-No, oTo nol v~ol~. prosody (as outlined in item 2 above) is common throughout his Italian operas and Notice that only at the beginning of the should be interpreted as a sign not that he fifth line does the secondary accent shift was careless with his text setting but rather from the second to the first syllable of the that he preferred a certain freedom in his line. (Example 7.) In his setting of this text approach to poetic stress. Clearly he did Verdi capitalizes on the consistent weak­ not wish to be locked into an immutable strong beginning of each line of verse, sat­ system of prosodic laws, especially when urating his melody with a corresponding he was trying to fit a particular text to a mythmic motive, melody which he had already composed.20 Another example from I Lombardi will serve to corroborate and expand these ob- to be found in Example 8.

WWe know from Verdi's sketches for and that he sometimes had whole Ian, 1941) pp. 64-65. See also the Introduction melodies in mind before ever seeing a word of to Martin Chusid's edition of Rigoletto in The the finished libretto he was going to be setting. Works of Giuseppe Verdi, Series I, Vol. 17 (Chi­ See Carlo Gatti, ed., Verdi nelle immagini (Mi- cago & Milan, 1983) p. xviii.

24 Example 8.

No! ho! jiu.s-ta. : ca.1.1 - Se. "11on e. o'H -: di -o fer -ra. :

Non non vo - +re ra. - .9e. in - i~ - ne. OU -tro.. - _9e. nest pa.5 )•ou.-

di Sa.n - s·J.i u.-,I -ma. -no far - pe.

l'eri - fer in -

...,." .5en·.so: p;- o che. ctll'o- ro ~- sta.-si de.I mon-s1.1.I-: '

Spi - re vo-tre. cle - Ii -re et le ,.,a.r - ty re. oe. )'non 0.. -

mo. - no . i~eJ -te. ct.I : cui. - lo 11or1 r pa.-: ro - It.

'a. vou..s lo.. hon -- te. Ci. vol.lb le cri - l'lle

This motive assumes such importance on the strong part of weak beats. Here it is as an integrative element in the cabaletta the fourth beat of each measure. that the composer refused to alter it to fit In example 6 we can see the French ver­ the one line of poetry in which the second­ sion of " Salve Maria" illustrated in paral­ ary accent did not fall on the second sylla­ lel with the original Italian. Primary poetic ble. What may be called Verdi's second op­ accents are fixed by the rules of French erative principle comes into play at the prosody, but secondary accents are deter­ point, m. 9, where the maintenance of mined by grammatical syntax. In example rhythmic symmetry becomes a more im­ 9 I have marked these syntactical cesuras in portant concern than exact poetic prosody. each line and located the secondary accents We also discover in this example that sec­ accordingly. Observe how closely this ondary accents frequently fall not only on scansion duplicates that of the original text. strong beats, as observed earlier, but also

25 Example 9. stress pattern oft 'empie il Signor. A glance

11 at example 6 will show this difficulty to Av~ Maria have been resolved in a purely accidental M'a voix I~ pri~ fashion. Since the original Italian setting was irregular with regard to stress, the mu­ iaris!mes pl~rs sic would have been perfectly regular for O vie~~ d~s d~le~rs . the setting in French: Note that the librettists have been careful in ~ t; _rf, m:s /pt rs almost every case to place the primary accent on the fourth syllable of each line. But lines 3 & 4 of the libretto (example 9) Only in the last line is the fourth syllable have been reversed in the music (example not the primary end stress. Thus, the pri­ 6) . Why? Certainly Verdi could have set mary accents of both the Italian and the the text as it was presented to him, and the French texts coordinate most of the time. resulting melody would have produced a Even the location of many of the secondary tolerably inaccurate French prosody. (Ex­ accents is identical. However, some diffi­ ample 10.) culty arises in the last two lines of the French where taris mes pleurs reverses the

Example 10. J J J J J J Ta-ris mes pleurs 0 Vier-ge des do- leurs

The reason for the switch of lines has to do here is simply a case in which he was able with the subject of stress and its most effec­ to fix some poor declamation without hav­ tive placement. The line that Verdi might ing to make melodic or rhythmic changes have written would have led to the major in his original musical line. But his priori­ musical accent at the climax of the phrase ties begin to show in other instances where falling on the relatively insignificant prepo­ his new French text did not quite fit the sition des. While French prosody may not earlier music. In those cases where the have been totally ruined by this slight irreg­ melody had been originally designed in ularity, Verdi obviously did not like the ef­ rhythmically symmetrical phrase groups, fect it would have produced. In reversing he refused to break the regularity of the the lines he managed to place the word des rhythmic pattern just to accomodate a in a much less conspicuous position while French accent. Refer back to example 8 for reserving the long, high accent for the an illustration from Helene's cabaletta word taris. This simple solution to a minor "Non, votre rage" in No. 3, Act ill, sc. 5 problem demonstrates that Verdi, even in of Jerusalem. This is the French version of his first , was well attuned to Giselda's "No giusta causa." The persis­ matters of stress in French texts. Does it tent weak-strong beginning of every me­ also mean that he was stricter or took more lodic phrase runs headlong into the word care with his French declamation than with votre (m . 6) in which the natural word his Italian? I think not. What we have seen stress falls on the first syllable. As in I

21In the printed libretto for the first performance (archive of the American Institute for Verdi line. Verdi 's autograph full score (film in Verdi Studies, film ML 49 , V 48, vol. 11, no . 3) we archive) shows "Ave Maria," as does the Escu­ find the words " Vierge Marie" in this opening dier piano-vocal score.

26 Lombardi, Verdi's setting ignores the nor­ effect than the similar placement of a femi­ mal prosodic values of the text by displac­ nine ending in an Italian word. ing the vo- of votre, a secondary accent, to One does find occasions when Verdi is the weakest position. In this particular in­ forced to change his original melody to ac­ stance the irregularity is compounded by comodate the differing stress patterns of the fact that mute endings in French po­ the new French text. In Pagano's aria etry, which are voiced very lightly, 22 are " Sciagurata" (I Lombardi, No. 2, Act I, much weaker than the regular piano end­ sc. 4) Verdi originally granted himself the ings of Italian words (e.g. amore). Thus, luxury of repeating a bit of text in the verse the placement of these " mute" syllables in "tu nel colmo, nel colmo del contento." relatively strong metric or rhythmic posi­ (Example 11. ) tions produces a more noticeably unnatural

Example 11 .

nel 'Co -11to ~el eol - 1110 ~I con-: fe11 -- +o I _,

~s - +e en - c.o - re. -- eo..-c'hl ll. lo.. fer - - re.

In his analysis of the transformation of I Kimbell implies that there is something Lombardi into Jerusalem, David Kimbell stiff or academic about the French version, refers to the French version of this aria whereas in fact , the new nine-syllable with the observation that "Verdi handles French line automatically precludes its set­ the French text with less freedom than the ting to the original eight-syllable melody Italian, allowing himself nothing of the im­ unless the composer is prepared to create passioned reduplication of words that one the kind of preposterous text setting illus­ finds at "tu nel colmo del contento. " 23 trated in example 12 .

Example 12.

.-3-, } P J }r f } ~ }. } J JJ J> Res-te... en- co-re en-co-re ca- c hr·a. l a ter-re

22 "As a resonance, an echo of the preceding syl­ lable, which vibrates, hovers, and gently dies in ard Strauss and Romain Rolland, p. 32. the air ... " Romain Rolland, as cited in Rich- 23Kimbell, p. 23 .

27 Thus, if there were a point at which mel­ 4) from Elisabeth's aria "Toi qui sus le ody took p-ecedence over natural prosody neant" {Act V, SC. 1) are typical. in Verdi's French text setting, so too there was a point beyond which the normal po­ Example 13. etic stresses of the language could not be distorted. That point seems to have been si l 'onJrep~d ~nc~~J~es larrrfs d~ns l~ fixed by Verdi at a limit of a single reversal ciel of a normal strong-weak or weak-strong syllable pair per line of poetry (in Italian) Por~Jn pleurantJmes pl~rsjla~x piectsJ or per hemistich (in French). The reason 1 example 12 is implausible is that this set­ d~ ·'E1~rn~1. ting reverses the normal prosody of both Line I is comprised of two identical hemis­ cache and la terre in the same line­ tiches that divide into groups of2+4 sylla­ something Verdi tends to avoid. bles. Line 2 is different. The first hemis­ tiche divides into 1+3+2 syllables and the m other into 2+4. This results in an AABA pattern to the internal division of these At this point one may ask if the special lines. Example 14 shows how Verdi chose circumstances surrounding the composition to set the first line to two identical two­ of Jerusalem might have produced proce­ measure phrases. But the rhythm of these dures for the treatment of stress different phrases is not determined by the prosody of from those found when Verdi worked from the text, as one might expect. Rather, the an original French libretto. To check this rhythm contradicts the prosody in three possibility we need only look at Verdi's last places: the initial weak syllable of each French opera, Don Carlos (1867). Here the hemistich is set on the strongest beat of the subject of musical prosody becomes more measure in both phrases; and in measure 4 complex. As in Jerusalem, much of the po­ the weak syllable dans also falls on the etry shows the French penchant for irregu­ downbeat, displacing the primary accent lar stress patterns. The following lines (3- ciel to a weaker metric position.

Example 14 .

Si ll>n ~-p4~~ en -c.o - re ~e.s- /o.1'·111es ~o.11S- It. ei1I

-I

While perhaps not as awkward as some of rhythmic-melodic symmetry. But in the the prosody we have seen, these four mea­ very next phrase he throws aside all con­ sures do represent a situation in which sideration of melodic symmetry and gives Verdi sacrifices normal prosody to the second line a much more natural musi-

28 cal prosody, creating in the process one. four-measure phrase with an asymmetrical, M'a foudrj;:j. 1e~s~/ / rorguei1 I non-repetitive rhythmic organization. Within a space of only eight measures des novate~rs. Verdi has shown us two extremes of me­ The pattern of stresses in the first two lodic style based on differing relative prior­ hemistiches is dissimilar, while that in the ities of text prosody and melodic shape. last two is similar. This means that hemis­ This variety of melodic style is not found tiches 3 & 4 could have been set as paral­ in Jerusalem and seems to indicate a new lel, congruent phrases. But Verdi's setting interest on Verdi's part in the creation of a (example 16) takes no notice of the similar more idiomatic prosody for the frequently stresses. Instead he sets these alexandrines irregular stress patterns of French poetry. as two asymmetrical three-measure I say "seems to indicate" because the re­ phrases, each of which begins with the lationship between melodic asymmetry and same rhythm but ends differently. We the proper setting of French texts with ir­ might be especially surprised by the musi­ regular prosody may not be of a cause-and­ cal congruence that Verdi forces onto the effect type, but may be coincidental. Sup­ first hemistiches of each line, for they are port for this view may be found .in other not even remotely alike in their prosody. Verdi melodies that are equally as irregular Here we have two hemistiches in line 2 that in their rhythmic or phrase structure as line have similar prosody but non-congruent 2 of example 14, but not because of the musical rhythms, and a hemistich at the be­ poetic prosody of the text. The ~ollowin? ginning of each line with dissimilar pros­ lines of Philippe from his duet with Rodn­ ody but congruent musical rhythms. This gue at the end of Act II (Tableau II, sc . 6) situation contradicts any theory propound­ contain two six-syllable hemistiches each. ing a cause-and-effect relationship betw~en irregular poetic stresses and asymmetncal Example 15 . rhythmic-melodic organization. J'ai /de ce prix 1s a'Ogl~t/!Paye /1~ p~ix v v du monde;

Example 16.

Philippe. D p v F r I F J"di de. ce. prix .sa.n-3fo.11t pa.-yi la. pa.ix ~u.. tltOll - de.

9·1h r Iqr a qp· i~ p· p I~IF F p· ~ F ,, Ma. fou.- are a. fer - ra.~ - .Se.. ll>r - 9u.eil dt.s 110 - VO. - +eu..rs

Budden talks about asymmetrical melodies Opera francese sulla struttura ritmica di like those of example 16 in his article ' Don Carlos.' " 24 He hypothesizes that " L'Influenza della tradizione de! grand Verdi was creating in Don Carlos a new

24In Atti del [/0 congresso intemazionale di studi verdiani. (Parma, 1971) 311 -18.

29 kind of liberated French melody, and at­ d'etre for this new liberated melodic style. tributes the new rhythmic freedom to the Verdi could easily have done the same­ suppleness of the French language and its and did according to Budden26 -with Ital­ "relative lack of tonic accent. " 25 The prob­ ian texts. lem with Budden's theory is that it con­ Looking back from Don Carlos to Jeru­ fuses suppleness with variety. Certainly salem, we see that the actual mechanics of French poetry contains a large variety of Verdi's French text setting did not change possible stress patterns, but this does not from his first to his last contact with that mean that the prosody of any particular line language. Rather, the procedures that is supple enough to be interpreted in sev­ guided him in the setting of French libret­ eral different ways. Yes, some unaccented tos were identical to those that governed syllables may, in their musical setting, his setting of Italian verse. They may be have to take on secondary accents in order summarized as follows: to avoid the impossibly long series of weak 1. Primary accents should fall on the syllables characteristic of French poetry. downbeat. But these secondary accents are deter­ 2. Secondary accents should normally mined by grammatical syntax; and since fall in metric positions stronger than this syntax is not something that can be re­ those occupied by the unaccented syl­ interpreted at will, the stress patterns of lables to which they are related. 27 French poetry are really just as fixed as 3. Exceptions generally occur when the those of Italian poetry. The difference is retention of rhythmic-melodic sym­ mainly that Italian stress patterns are fixed metry is critical for the musical­ before the fact, so to speak, by poetic me­ dramatic character of a number. In ters, while French stress patterns are fixed other words, musical considerations after the fact by syntactical relationships. sometimes take priority over textual In any case, we can find nothing inherent considerations. in the French language, per se, that allows 4. Normal prosody tends not to be dis­ Verdi to set the first half of each line in torted more than once in a line of Ital­ example 15 in rhythmic parallel. If one in­ ian poetry or in a hemistich of French vokes, as does Budden, the relative weak­ poetry. ness of secondary accents in French poetry As his letters to Du Lode, quoted earlier, as a justification for situations like these, show, Verdi insisted that some of the po­ then one would have to admit the possibil­ etry of Don Carlos be constructed with an ity of any kind of metric placement for sec­ Italianate regularity to accomodate a par­ ondary accents. But as we have seen, Ver­ ticular melodic idea he apparently already di's treatment of these accents is, on the had in mind. Clearly these were not the de­ whole, more predictable and methodical sires of a man who felt his music had to than this and gives no indication of having conform to every nuance of the French lan­ been governed by such prosodic license. In guage. Rather, the creation of an effective short, we cannot use language as the raison musical prosody became for Verdi a deli-

25/bid. p. 313 . 26Budden contends that Verdi adopted this new are part of a feminine word ending, in which sty le in his later Italian operas after having cre­ case they are related to the accent immediately ated it under the influence of the French libretto preceding. The most common exception to prin­ for Don Carlos. ciple no. 2 occurs when the unaccented syllable 27Syllabic relatedness may be defined as follows: of a feminine ending after a primary accent falls unaccented syllables are related to the following on a downbeat without upsetting its normally accent except when those unaccented syllables weak feeling.

30 cate process of balancing the often conflict­ librettos, and that changes of melodic style ing demands of poetic stress with the exis­ in Don Carlos cannot be attributed to these tence of independently conceived melodic procedures, should be sufficient proof that structures. The fact that this balance was Verdi's French text setting was not signifi­ achieved using the same system of operat­ cantly affected by any unique characteris­ ing procedures for both French and Italian tics of that language.

31