Style in Singing

PORTAMENTO

THIS is effected by the voice gliding from one tone to another, and is equally available on stringed instruments, the violin or ’cello, the mandoline or zither. It is a grace of style much abused by inartistic singers. Being an ornament, good taste dictates that it be used sparingly. A frequent sliding from one tone to another is a grave fault, and most disagreeable to a cultivated ear. To sing legato is one thing; to sing strisciato is another. Hence, its use on two consecutive occasions is rarely admissible. But without a sober and discreet use of the portamento, the style of the singer appears stiff, angular—lacking, as it were, in graceful curves.

It must always be performed by carrying the tone and syllable to the next tone; never by anticipating the latter:

[Listen]

But it sometimes happens that, while desiring this grace, the composer does not indicate his wish quite correctly. Here is an instance by F. Thomé:

[Listen]

Were it performed as printed, it would be very bad style, as it violates the rule that the succeeding syllable shall not be anticipated. Undoubtedly, what the author wished is the following:

[Listen]

Sometimes the composer himself indicates clearly his intention that this effect should be used, as in the following examples:

[Listen]

[Listen]

[Listen]

[Listen]

(Notice the phrases marked a and b.)

The words and indications for the use of the portamento in each of these last four examples are by the respective composers, and as printed in the published editions.

A portamento should never be sung so slowly as to convey the idea of a badly executed chromatic scale; and, as a rule, it is best not to use one between any lesser interval than a third, unless for some particular effect, or at the close of a slow movement, as in the aria “He was despisèd,” in The Messiah:

[Listen]

It is also effective in connecting syllables in phrases of a smooth, lyric character:

[Listen]

The portamento being an embellishment that pertains to the cantabile, it is very little used in declamatory singing.

But frequently in the Recitatives of classic works occur phrases of declamatory recitative, interspersed with passages that are purely lyric in structure. To each of these divisions must be given its appropriate style. For instance, after the opening phrases of Obadiah’s exhortation, “Ye people, rend your hearts,” in Elijah, up to the end of the phrase “Return to God,” all is purely lyric declamation. But at the words, “For He is slow to anger, and merciful,” this should cease, and the succeeding phrases be given with all the graces that are permissible in cantabile singing; not in the hard, dry manner affected by some of the modern tenors in oratorio.

[Listen]

VARIATIONS OF TEMPO

THESE are of value in bringing out the musical and poetic significance of certain compositions; notably the operas of Bellini, Donizetti, and the earlier works of Verdi. But I would caution singers to exercise discretion in this much-abused effect. Variations of Tempo, the ritardando, accelerando, and tempo rubato, are all legitimate aids demanded by Expression. But unless their use is determined by sound judgment and correct musicianly taste, the effect speedily becomes vulgar and monotonous. Knowledge, and a taste formed in good schools, must be the guide of the vocalist in the use of variations of tempo.

I have said that the operas of Bellini, Donizetti and Verdi abound in instances requiring the hastening or slackening of the tempo. But the device is also highly esteemed by the ultra-modern Italian school, as may be seen in studying the scores of Puccini, Mascagni and Leoncavallo.

Here is an illustration of its effective use in the air “Connais-tu le pays?” from Mignon (Act II), by Ambroise Thomas. Madame Christine Nilsson (Countess Casa Miranda), who “passed” the rôle with the composer, always sang the phrase thus, although these indications do not appear in the published version:

[Listen]

Again, in the fine song Der Asra, by Rubinstein, the musical, as well as the dramatic, effect of the poem is heightened by the use of the accelerando, which interprets with musical vividness the impetuous avowal by the slave of his passion for the princess, after his calm answer to her questions as to his name and birthplace.

“Ich heisse Mahomet, ich bin aus Yemen, und mein Stamm sind jene Asra, welche sterben, wenn sie lieben.” (Heine.)

[Listen]

CHAPTER IV

Tradition

In the original editions of many of the earlier operas, as those of Mozart, etc., the unaccompanied recitative (recitativo secco) is not barred. As with the plain-chant of the church, only the pitch of the tone is indicated. Its length was left to the discretion of the artist, who was supposed to be familiar with the accepted style of delivery termed “recitativo parlante.” The example is from the recitative “Dove sono,” in Act III of Le Nozze di Figaro, by Mozart:

[Listen]

This should be sung as below:

[Listen]

The substitution of another note for the one actually written, both in Recitative and Aria, was also strictly regulated under the system or convention then in vogue, one perfectly understood both by composer and singer.

In all the earlier Italian operas, and in the English oratorios of Handel, this system was followed:

[Listen]

[Listen]

[Listen]

[Listen]

This substitution, therefore, of another note—a tone or semitone higher or lower, according to the phrase—is not only legitimate but essential in all music written in the Italian manner.

Another cause of changes being necessary in the vocal part of many of the older classic writers, particularly of oratorio, is the frequently faulty syllabic accentuation. I have already mentioned this defect in the chapter on Accent. Handel, for instance, although living nearly all his life in England, never became quite master of its language; hence the numerous cases of the misplacing of syllables in his oratorios. This defect is also noticeable, but not in the same degree, in his Italian operas. The books of Elijah and St. Paul (Mendelssohn), and The Creation (Haydn), were originally written in German, and therefore suffer somewhat in this respect when the translated English version is given. This fault is also noticeable in the English versions of Bach’s Passion (St. Matthew), and Mendelssohn’s Psalm CXIV. In the first quoted of these two works, in the response for Double Chorus to the question, “Whether of the twain will ye that I release unto you?” the accent falls on the first syllable “Ba-rab-bas”; in the second of the two works (114th Psalm), the accent is placed on the last syllable, thus: “Hal-le-lu-jah.” Neither of these accentuations is in accordance with English custom.

A singer, therefore, is perfectly justified in rearranging the syllables in order that, as far as possible, the musical and verbal accents shall coincide. But there are rigorists, unaware of the usages and conventions previously spoken of, who are very severe in their judgment when any deviation is made from the printed score with which they follow the performance of classic works. Such severity is unmerited, because unjust. Although such persons sometimes inveigh against any and every change from the strict letter of the printed music—ignorant of the possibility, that only in this way can its spirit be respected—the changes in a multitude of cases are essential because due (1) to reverential deciphering of an obsolete musical notation, (2) to improvements in musical instruments, or (3) to the sanction and authority of the composer himself.

Sometimes it is an orchestral conductor who reproaches the solo singers with their want of respect for the composer, because he hears at times interpolations or changes which find no place in his own score. The singers are accused of “altering the composer,” of “taking liberties with the text.” And yet these very changes may be traditionally correct; they may be in accordance with rules and conditions prevalent at the time the music was written, and employed on account of a desire to interpret the composer’s own intentions, and not from mere vanity or caprice.

Nor are these necessary changes and departures from the printed scores of the classics confined to the vocal parts of the music composed by the old masters. As a matter of fact, the deviations which, in performance, are sometimes made from the printed edition of a musical composition, arise from a variety of causes.

One of these is the discrepancy that exists between various editions of the same work; and sometimes the confusion is complicated by different versions having been prepared by the composer himself. This is notably the case with Gluck’s Orphée, first written to an Italian libretto by Calzabigi and produced at Vienna. When Marie Antoinette called her former Viennese singing-master, Gluck, to Paris, she gave him an opportunity of displaying his genius by facilitating the production of his Iphigénie en Aulide at the Opéra, in 1774. Its enthusiastic reception recalled to the composer the like success which had attended the production of his Orfeo at Vienna. He immediately set to work to revise it for the Paris Opéra, and fit it to a new French text, the latter supplied him by Moline.[2]

But the title-rôle in the original Italian version was written for, and sung by, Guadagni, an artificial contralto (contralto musico). In its newer French dress the part was transposed and rearranged for the tenor Legros; who, judging from the extreme altitude of the tessitura employed, must have possessed either a haute-contre, or a very high light-tenor voice, and who may have employed the falsetto. This high tessitura, combined with the fact that the pitch has risen considerably since it was composed, renders the French version impracticable for tenors of the present day. Here are the concluding bars of the famous air as written in the original Italian version, and the same phrase as altered by Gluck, when produced in Paris.

(As originally written by Gluck for the Italian version, Vienna.)

[Listen]

The finale to the first act was also changed; a tumultuous “hurry” for strings, evidently designed to accompany the change of scene to Hades, being now replaced by a florid air, probably introduced at the desire of the principal singer as a medium for the display of his vocal virtuosity; a concession often exacted from composers of opera. This interpolated air was for a long time attributed to a composer—Bertoni—who had himself composed an opera on the subject of Orphée. Later researches have, however, proved that this air is by Gluck himself, taken from Aristeo, one of his earlier works. When the famous revival of Orphée took place at the old Théâtre-Lyrique in Paris, the rôle of Orphée was restored to the type of voice—contralto—for which it was originally composed, and confided to Mme. Pauline Viardot-Garcia. She retained the air introduced for the tenor Legros, but of course transposed, and with a reorchestration by Camille Saint-Saëns; the now famous composer having at that time, by the request of Berlioz, undertaken to continue and complete the revision of Gluck’s complete works, known as the Pelletan Edition.[3]

Other changes from the first Italian score were also made by Gluck in the later French version. Here is an example; being the recitative immediately preceding the great air of Orpheus in the last act:

(Original Italian version, as written for Vienna.)

[Listen]

(As written for the Paris version, the rôle of Orphée being then sung by a tenor.)

[Listen]

Again, discrepancies exist between various published copies of the same work, arising from the fact that sometimes the editors of these revisions may have mistaken the intentions of the composer. Or, influenced by pardonable human vanity, they may have felt impelled to collaborate more directly with the composer, by adding something of their own.

There is valid reason for the additional accompaniments, with which Mozart has enriched the original scores of Handel’s Messiah and Alexander’s Feast; and we have evidence of the skill, and can divine the reverence, with which these additions were accomplished. But how fatal would have been the results, had the delicate task been attempted by one in whom these qualities were lacking! Also, there is every excuse for the additions made to Gluck’s Armide by Meyerbeer for the Opera of Berlin; and we have the direct testimony of Saint-Saëns, who has examined this rescoring, as to the rare ability and artistic discretion with which the work has been done.[4]

From this evidence it appears that in the score as left by Gluck, the trombones do not appear at all in Armide. The drums, and stranger still, the flutes, are heard only at rare intervals; while the whole orchestration—sometimes a pale sketch of the composer’s intentions—shows a haste and lack of care in marked contrast with the pains bestowed on the scoring of Alceste, Iphigénie, and Orphée. The revisions and additions spoken of were undertaken by highly competent authorities, actuated only by the wish to restore in its purity the idea of the composer; and who to zeal, added the more valuable quality of discretion.

Ancient music, owing to the development of and changes in the instruments for which it was composed, can rarely be given as written by the author. Even if the instruments of modern invention be eliminated, the orchestra of to-day is not the orchestra of Handel. The oboe, for example, has so gained in penetrating power that one instrument to each part now suffices; in Handel’s time the feeble tone of the oboe rendered a considerable number necessary. The perfection of certain instruments, too, is the cause of modifications in the music written for them. The limited compass of the pianoforte, for example, was certainly the sole reason why Beethoven failed to continue in octaves the entire ascending scale in one of his sonatas. Had the piano in his day possessed its present compass, he would undoubtedly have written the passage throughout in octaves, i.e., as modern pianists play it. If a rigid adherence to the printed letter of ancient music is to be strictly observed, without consideration of the many causes that render this procedure undesirable, let consistency be observed by pushing the argument to its logical conclusion, viz., returning to the instruments used, and the composition of the orchestra that obtained, when these works were written. Those who accuse artists of introducing changes, of not performing the music as the composer wrote it, should be quite sure as to what the composer really did write, since many changes are made both before and after the work is printed. They should also be certain that these changes are not such as the composer may have, or would have, sanctioned, seeing that by their use his meaning is more clearly expressed.

At the Concerts Spirituels, given at the Church of the Sorbonne, Paris, may be heard very excellent performances of Oratorio by ancient and modern composers, from Handel and Bach to Claude Debussy; though I do not know whether or no l’Enfant prodigue (The Prodigal Son), by Debussy, is properly styled an oratorio, seeing that it was recently given in London on the stage as an opera. These performances at the Sorbonne are marked by a reverential attention to detail; the soloists, chorus and orchestra being very competent, and the conductor—M. Paul de Saunières—a musician of ability and experience. In spite of these great advantages, however, the works of several of the old classic composers suffer somewhat, by certain authentic traditions and conventions being either unknown or ignored. To cite only one instance out of many: At the Sorbonne, the opening bars of the second movement of the Recit. in The Messiah, “Comfort ye my people,” etc., are performed as printed:

[Listen]

This music is written in the Italian “manner,” consequently its performance should be in conformity with the usages and conventions which obtained when the work was composed. One of these, as I have pointed out, was the substitution of one note for another in certain places; another, that in declamatory recitative, or recitativo parlante, the chord in the orchestra should come after the voice (“dopo la parola”). These words appear in many scores of the Italian operas, even of the present day. But when they do not, the musical director is supposed to be familiar with the custom. The following, therefore, is the authentic mode of performing the passage in question:

[Listen]

Apart from these defects in the rendering of the ancient classics, it would be unjust not to acknowledge the great artistic merit and value of the performances, given—as Oratorio should be—in the church. To hear l’Enfance du Christ (Berlioz) as performed at the Sorbonne, with its particular facilities for obtaining the ppp effects of the distant or receding angelic chorus, is to be impressed to a degree impossible of attainment in the concert-room.

Let those purists who resent any “tampering”—as they term it—with the composers’ music listen to the following phrase, sung as it is printed in the ordinary editions:

[Listen]

Then let them hear it given according to the authentic and accepted tradition, and say which of the two versions most faithfully interprets the composer’s meaning.

[Listen]

Let us now consider alterations which do not appear in the printed editions, and yet may have been made or sanctioned by the composer.

In comparison with painting and sculpture, music and the literature of the theatre are not self-sufficing arts. They require an interpreter. Before a dramatic work can exist completely, scenery, and actors to give it voice and gesture, are necessary; before music can be anything more than hieroglyphics, the signs must be transmuted into sound by singers or instrumentalists. Wagner embodied this truth in his pathetic reference to Lohengrin: “When ill, miserable and despairing, I sat brooding over my fate, my eye fell on the score of my Lohengrin, which I had totally forgotten. Suddenly I felt something like compassion lest the music might never sound from off the death-pale paper.” In other words, Lohengrin, though finished in every detail, was merely potential music. To make it anything more, the aid of singers and orchestra are essential.

Composers and dramatic authors, in fact, create their art-works; but it is their interpreters—actors, singers, instrumentalists—who animate them, who breathe life into them. One of the inevitable consequences is, that the composer’s ideal can never be fully attained.

But changes in performance from the printed text of a composition are frequently the work of the composer himself. If really an artist, he is rarely perfectly satisfied with his completed work. The difference between his ideal and his materialization of it, is a source of anguish for him. The journey made by a vision of art from the brain that conceives it to the hand that imprisons it in marble, or depicts it in colour, or pens it in words or music, is a long one. And much grace or power, beauty or grandeur, is inevitably lost on the way. This is the explanation of the disappointment of all true artists with their creations. This is the origin of their endless strivings to perfect their works; the first embodiment is not a perfect interpretation of the artist’s inspiration, and further reflection has revealed to him an improvement. The process is endless.

If one wishes to surprise genius labouring to give birth to perfection, one should consult the later editions of Victor Hugo’s works and note the countless emendations he made after their first publication—here a more fitting word substituted, there a line recast, elsewhere an entire verse added, or excised, or remodelled.

This work of incessant revision is not restricted to poets. Composers of genius are also inveterate strivers after perfection, are continually occupied in polishing and revising their music. And not all the modifications they make, or sanction, are recorded in the printed versions. For many are the outcome of after-thoughts, of ideas suggested during the process of what I have called transmuting musical hieroglyphics into sound. Such modifications, usually decided upon in the course of a rehearsal—I am now considering particularly operatic works—are frequently jotted down, a mere scanty memorandum, on the singer’s part or the conductor’s score. But they are the work of the composer, or have received his approval, and, although not noted in the printed editions of his compositions, are transmitted orally from conductor to conductor, singer to singer, master to pupil. And thus a tradition is perpetuated.

But the question of changes goes even further.

Prior to the advent of Wagner, the singer was allowed great license in operatic works. This license was principally manifested in a two-fold form. The first is called pointage (French), puntatura (Italian), and means the changing of the notes or contour of a musical phrase; the second is termed changements or variantes (Fr.), abbellimenti or fioriture (It.), and refers to the interpolation and addition of ornaments, i.e., embellishments and cadenzas.

POINTAGE

[Listen]

Every vocalist knows the difficulty experienced in singing very high tones to different syllables, each requiring a different conformation of the buccal cavity. The passage quoted—expressing Tell’s bitterness at the recollection of his past sufferings in prison, “Well I know the weight of galling chain”—has to be declaimed with great energy. So far as the relative value of the notes is concerned, it is entirely ad libitum, the rhythmical figure in the orchestra having ceased one half-bar before. It is said that Dabadie, a basso cantante rather than baritone, to whom was entrusted the rôle of Tell on the first production of the work at the Opéra, Paris, on August 3, 1829, finding it impossible to sing the phrase as written, had recourse to a professor. He advised the pointage given later. This change became traditional, and has since been followed, except, it is said, in the case of Massol, who succeeded Dabadie. He, being possessed of a very sonorous voice of exceptional compass, was able to give the phrase as written. This change, or pointage, must have been heard by Rossini, and so must have been tacitly approved by him. This is the change made by Dabadie:

[Listen]

In Italian lyric theatres, pointage becomes necessary in many French operas, owing to the prevalent custom of allotting to contraltos certain rôles written for soprano and known as “dugazon rôles” (from Madame Dugazon, who created the type). The parts of Siebel in Faust (Gounod), Urbain in Les Huguenots, Stéphane in Roméo et Juliette (Gounod), are all written for soprano, and when sung in Italian require not only transposition of the principal airs, but the use of pointage in passages where transposition is impossible owing, for instance, to the participation of other characters in the scene. Thus the air sung by the page Urbain (Les Huguenots) on his entrance is sung in the French theatres as written by Meyerbeer, i.e., in B flat. In theatres where the Italian version is given, this air is transposed a third lower into G, necessitating later numerous pointages, for the reason already given.

I said that many deviations from the printed text are the work of the author, or are authorized by him. A moment’s reflection will convince one of the truth of this statement. The singer chosen—usually by the composer himself—to “create” a rôle, i.e., to interpret for the first time some part in a new opera, generally studies it with the composer, or under his direct supervision, and thus learns, directly or indirectly, his ideas as to the meaning, style of execution, tempi, etc., of the music. Very often during rehearsals, when the composer begins really to hear his own work, he makes modifications in certain passages, alterations of the words or suppressions of the notes that are either ineffective, or lie awkwardly for the voice. But the opera has already been printed for the convenience of the singers and choristers studying the rôles and choruses; consequently, such modifications, rearrangements, and “cuts” (as excisions are termed), do not find their way into the published scores.

Meyerbeer, as I have been informed by competent authorities, was constantly modifying his compositions. With him, the work of revision and emendation was never finished. It is said that this was more especially the case with his last opera, l’Africaine, which he was continually altering and revising, never being able to satisfy himself. Two versions of the libretto were prepared for him by Scribe, and two distinct settings of the music are published, although only one is performed.[5]

In Nelusko’s first air occurs the following passage, in which a great crescendo is marked, culminating ff on the word rien:

[Listen]

Although the opera was produced after the composer’s death, Jean-Baptiste Faure, the great baritone chosen to create the rôle of Nelusko, studied it with Meyerbeer, who authorized several verbal and musical changes in it.

[Listen]

Without the first alteration it is impossible to realize the composer’s wish for a climax on the word “rien”; the second change is due to the fact that the tessitura of the phrase is somewhat high, and Faure, who was a low rather than high baritone, dreaded the high f♯.

Indeed, it was for this latter reason that this most accomplished singer never sang in Verdi’s operas. According to his own statement, he had to deny himself this pleasure, because most of the baritone parts in the Italian composer’s operas are written in a high tessitura.

When Gounod wrote his Faust for the Théâtre-Lyrique, Paris, spoken dialogue was used in place of the recitatives subsequently added by the composer when the work passed, ten years later, into the répertoire of the Opéra. In its earlier form, therefore, it belonged to the category of opéra-comique, in which tenors were then permitted to use the falsetto voice for their very highest tones. This custom, though sanctioned in opéra-comique, was not permitted or accepted in grand opéra, to which Gounod’s work in the revised form now belongs. At the beginning of the sixth bar from the end of the tenor cavatina in the Garden Scene: “Salut! demeure chaste et pure,” occurs the high sustained c.

Not all tenors who sing the rôle are possessed of the much-coveted “do di petto,” so a discreet pointage becomes a necessity, since the tone was originally intended, as I have said, to be sung in falsetto. Those robust tenors who, possessing this tone, launch it out at full voice, unheeding the delicate accompaniment with violin obbligato in the orchestra, and the calm, mystic serenity of the surroundings, are surely more desirous of drawing the attention of the public to themselves, than actuated by an artistic desire to interpret faithfully the scene as intended by composer and librettist.

It was owing to the use by light tenors of the so-called falsetto voice, now no longer in favor with the public, that such of the opéras-comiques by Boiëldieu, Halévy, Auber, etc., which still keep the stage, necessitate frequent pointage, in order to render their execution compatible with existing requirements. Sometimes a composer utilizes an exceptional voice, as was the case with the rôles written for Martin. This singer must have possessed either a strong tenor voice with exceptional low tones, or a baritone voice with perhaps an unusual command of the falsetto—history furnishes but vague information on this point. In any case, the rôles written for him—called Martin-tenor or Martin-baritone parts—are now assigned to the ordinary baritone. Pointage then becomes inevitable, as in the case of Hérold’s Zampa, the compass required as printed being from

[Listen]

In the rôles, such as Mignon (Thomas) and Carmen (Bizet), written for Madame Galli-Marié, their respective composers themselves have so arranged the parts that they may be sung by either mezzo-soprano or soprano. The rôle of Mignon has alternatives, in order that it may be sung by three types of female voices. The roulades and cadenzas were subsequently added by the composer for Madame Christine Nilsson.

If the rôle is sung by a high soprano, Mignon’s first air, “Connais-tu le pays,” is transposed a tone higher into E flat.

In the famous duet between Raoûl and Valentine in the fourth act of Les Huguenots, the composer has given alternative notes for those tenors who do not possess the exceptional altitude required for the higher of the two:

[Listen]

I heard recently, however, a performance of this opera, in which the tenor sang the whole of the music as written, without either transposition or pointage. So it was sung, I should imagine, by the famous Adolphe Nourrit, who created the rôle; but the pitch at that time (1836) was lower than it is at present.

Thus composers have recognized the necessity at times of pointage in certain rôles written for exceptionally gifted singers, in order to render possible to the many that which was originally written for the few.

Changes from the published version have also been made—and proving effective have passed into tradition—by singers who, exercising the liberty then accorded them by composers, have slightly modified certain passages for several reasons: for instance, to augment the effect by making the phrase more characteristic of the vocal instrument, or to express more forcibly the composer’s idea.

The following illustrations will render my meaning clearer. The changes originated in the causes I have mentioned, and are attributed to Madame Dorus-Gras:

[Listen]

The phrase “Grâce, grâce,” in which Isabelle implores Robert of Normandy’s forgiveness, occurs three times. When it recurs for the last time, a change from the printed text is not only justifiable; it is demanded, in order to give additional intensity and power to the phrase, and to avoid the monotony caused by mere repetition. This modification is all the more defensible, as the composer has substituted the orchestra, with the strings tremolo, for the rhythmical harp-figure with which he accompanies the phrase on its first and second presentations. Here is the accepted traditional change:

[Listen]

Again, to sing the final cadenza of this air as Meyerbeer briefly indicated it, would be impossible and absurd:

[Listen]

Other changes have their origin in the fact that sometimes a great climax is rendered impossible of realization because the musical phrase culminates on a vowel-sound difficult of emission on that note, and devoid of sonority; another word has sometimes to be substituted. For this reason, in the first air of Alice in the same opera (Robert), “Va, dit-elle,” a verbal rearrangement is always resorted to:

[Listen]

To avoid the disagreeable and ineffective result produced by the high descending passage on the word “lui” (pronounced in English as “lwee”), the last few bars are performed thus:

[Listen]

When La Tosca (Puccini) was produced in French at the Opéra-Comique, Paris, the unfortunate artist to whom was allotted the tenor rôle was expected by the translator to sing at full voice, and after a crashing chord from the entire orchestra, marked ffff in the score, the following words:

[Listen]

As it was found to be out of the question to produce the effect desired with the words as they stood, the phrase was afterwards changed to:

[Listen]

Frequently modifications, most happy in their effect, are due to the inspiration of a particularly gifted artist.

Madame Viardot-Garcia, finding the phrase of the cabaletta in the aria “Se Romeo t’uccise” (Romeo e Giulietta, Bellini) somewhat weak and ineffective, made the skilful pointage here given:

[Listen]

A great artist may feel at times the inadequacy of the phrase as it stands to convey justly the composer’s idea. Take, for instance, the well-known change which every soprano who sings the rôle of Leonora introduces in the Miserere scene of Il Trovatore. The passage occurs four times in succession, and as printed becomes commonplace and monotonous.

[Listen]

The accepted traditional change certainly conveys the impression of Leonora’s gradually increasing anguish and terror; not the idea that it is introduced merely to exploit a high tone:

[Listen]

That this departure from the text must have been sanctioned by Verdi, is, I think, proved by the fact that it has always been sung thus, and the composer himself must often have heard the substitution. He would certainly have forbidden its use, had he not approved of it, for he was particularly averse to having changes made in his music. The following anecdote illustrates this trait in his character. It was related by the late Mme. Marie Saxe, better known under her Italianized name of Marie Sasse. This distinguished soprano singer, a member of the Paris Opéra for a number of years, was engaged to give a certain number of performances at the Opera of Cairo. Aida was one of the operas stipulated for in her contract. She had never sung the rôle, and in studying it found the tessitura of the music, at one or two points, a little too high for her natural means. As she was compelled by her contract to sing the opera, she asked Verdi to make some slight changes to bring the music within her reach. But he refused absolutely to make the least alteration.

Madame Saxe was specially selected by Meyerbeer to create the rôle of Sélika in l’Africaine. She studied the part for three months with the composer, and sang it when the work was first given at the Paris Opéra. She was also chosen by Richard Wagner for the part of Elisabeth when Tannhäuser was given its stormy performances, with Niemann in the title-rôle, at the same theatre in 1861.

Madame Saxe possessed a score of Tannhäuser with the inscription in the composer’s handwriting: