The ordinary opera-goer is likely to treat the libretto with calm contempt. He is prone to assert that nobody cares about the words, and he does not reflect that behind and beneath the words is the supporting structure of the story. After all, an opera is a play, it is a music-drama, and the plot is as important in a play the words of which are to be sung as in a play the words of which are to be spoken. True it is, of course, that in an opera the words may not be heard distinctly, and perhaps they need not be seized with certainty, since the emotion they set forth is more amply conveyed by the music. But the musician cannot express emotion musically, unless there is emotion for him to express, unless he has characters immeshed in a series of situations which evoke vivid and contrasting sentiments for him to translate into music. As the music-drama is a drama, it must obey the laws of the drama; it must represent a conflict of contending desires; it must be carried on by characters firm of purpose and resolute in achieving their several aims. These characters must be sharply individualized and boldly contrasted; and the story in which they take part must be at once strong and simple, calling for no elaborate explanation and moving forward steadily and irresistibly. It must have a lyric aspect, lending itself naturally to song; and it ought also to afford opportunity for the spectacular effects appropriate to the large stage of the opera-house.
So contemptuous of the libretto is the ordinary opera-goer that he rarely inquires as to the name of the author of the book, altho he is generally familiar with the name of the composer of the score. He may or he may not be aware that Wagner was his own librettist, and quite possibly he supposes that it is the ordinary custom of the composers to write the words for their own music. He knows that 'Carmen' was composed by Bizet, and that the 'Huguenots' was composed by Meyerbeer; but he would be greatly puzzled if he was asked to name the librettists of these two operas, the adroit playwrights who devised the skeletons of dramatic action which sustained the composers and provided them with ample opportunities for the exercise of their melodic gift. As a matter of fact, the book of 'Carmen' was written in collaboration by two of the most distinguished French dramatists of the nineteenth century, Meilhac and Halévy, the authors of 'Froufrou' and of the librettos of Offenbach's 'Belle Hélène,' 'Grand Duchess of Gérolstein,' and 'Périchole.' And the book of the 'Huguenots' was the work of the master stage-craftsman, Scribe, the author of 'Adrienne Lecouvreur' and of the 'Ladies' Battle,' and of countless other plays performed in every modern language, and in all the countries of the world.
Bizet wrote other operas besides 'Carmen,' and if these other operas have vanished from the stage, the reason may be that the librettos to which they were composed were not as ingenious and not as interesting as the book of 'Carmen.' One of these forgotten operas of Bizet's was a dramatization of the 'Fair Maid of Perth,' and another was called the 'Pearl Fisher'; but neither of these books was devised by Meilhac and Halévy. And Scribe was not only the librettist of the 'Huguenots' and of the 'Africaine' for Meyerbeer; he also wrote the books of 'Fra Diavolo' and of 'Crown Diamonds' for Auber, the book of the 'Dame Blanche' for Boïeldieu, and the book of the 'Juive' for Halévy. Indeed, it is evident that Wagner himself as a librettist must be considered as a direct disciple of Scribe; certainly his book of the 'Flying Dutchman' has its points of resemblance with the books Scribe invented for 'Robert the Devil,' and for the 'Prophet.' Even the libretto of Wagner's 'Master-Singers of Nuremberg,' altho it is far richer in tone than any of Scribe's librettos for Auber, is constructed in accord with principles already applied by the French playwright. In fact, the influence of Scribe is patent thruout the long history of opera in the nineteenth century; he was not only the most prolific of librettists himself, but the operatic formula he devised was borrowed by the best of the librettists who followed him. Scribe was not the writer of the books of 'Faust,' or of 'Roméo et Juliet,' or of 'Aïda,' but all these librettos were carefully built in accord with the principles that he had practised for half a century.
II
Probably the average opera-goer is contemptuous of the libretto, because he thinks it is an easy task to write the mere words of an opera. To him, no doubt, the opera lives by its music, and by its music alone. But there is really no warrant for this uncomplimentary attitude. An opera is a music-drama, and if it is to achieve success, wide-spread and long-lasting, its drama must be as effective as its music. Experience proves that, so far from being as easy as it seems, the construction of a satisfactory libretto is really a difficult feat, to be achieved only by an expert in stage-craft. It is no task to be confided to an amateur play-maker, to a mere lyrist, ignorant of the art of the theater. First of all, a satisfactory book must contain the skeleton of a good play; and, second, this must be the special kind of play which will not only inspire the musician, but afford him a succession of special opportunities for the exercise of his own art. The book of an opera must be a good play; and more than once have we seen a libretto deprived of its music and written out again in prose for production in non-musical theaters. 'Carmen' is one example of this transformation. The late Sir Henry Irving was so taken with Wagner's 'Flying Dutchman' that he had it made over into a play for his own acting—'Vanderdecken.'
The book of an opera must be a good play, and therefore not a few successful operas have been composed on plots which had already won approval as plays on the stage. Indeed, many modern composers are so convinced of the necessity that librettos shall be attractive in themselves that they are continually borrowing popular plays to deck with melody. 'Salomé' and 'Pelléas et Mélisande,' 'Madam Butterfly' and 'Cavalleria Rusticana,' the 'Bohème' and the 'Tosca' were all successful without music before they were set to music to win a second success. The book of Verdi's 'Rigoletto' is based on Victor Hugo's drama, 'Le Roi s'Amuse'; and oddly enough it was the operatic libretto, rather than the original poetic drama, which suggested the English play on the same theme, Tom Taylor's blank-verse drama, the 'Fool's Revenge.' Another of Verdi's librettos was borrowed from Hugo's 'Hermani', while his 'Traviata,' as we all know, is taken from the play of the younger Dumas, long popular in America as 'Camille.' Two of Verdi's latest operas had Shaksperian themes, 'Otello' and 'Falstaff.'
It is instructive to note, so an American musical critic once asserted, that of all Gounod's dozen operas, "the only two which have survived are the two which are derived from Goethe's 'Faust' and from Shakspere's 'Romeo and Juliet'"; and he added a reminder that in these operas the music owes its success "not only to the aid derived from its associations with a favorite play, but also in part to the fact that the composer's creative imagination was fertilized by the splendid opportunities for dramatic composition offered by these plays. Gounod was moved by the joys and woes of Margaret and of Juliet, and it is only under the influence of deep feeling that such masterworks can be created." When Gounod set to music a poetic play by Goethe, and when Verdi set to music a group of characters created by Shakspere, the composers might well be inspired by the poets; and they were thus aided to attain the utmost of which they are capable as musicians.
But it may be doubted whether any musician could find any really helpful inspiration in dramas of vulgar violence, such as the 'Tosca' of Sardou, and the 'Salomé' of Oscar Wilde; and it is extremely improbable that the operas composed to such unworthy themes will be able to achieve any durable popularity. In plots of so coarse a character there is neither beauty nor poetry, and the vogue of music-dramas having subjects so debased is likely to be fleeting. On the other hand, there was both poetry and beauty in the original plays of 'Madam Butterfly' and 'Cavalleria Rusticana,' and we need not be surprised if the operas composed on these themes prove to have a long life in the musical theaters. We may even go further and suggest that there was a haunting and ethereal grace about Maeterlinck's 'Pelléas et Mélisande' which seemed almost to demand translation into the sister art of music.
The two most effective French comedies of the eighteenth century, the 'Barber of Seville' and the 'Marriage of Figaro,' supplied librettos, one for Rossini and the other for Mozart. We may be sure that sooner or later some other composer, Italian or American or German, will be tempted to undertake an opera based on Fulda's 'Two Sisters,' in which there could not help being a very effective part for the prima donna. And sooner or later again some musician with an appreciation of humor and sentiment will be moved to take for his libretto the comedy of 'Masks and Faces,' by Charles Reade and Tom Taylor, generally known by the name of its fascinating heroine, Peg Woffington. No doubt there are not a few other modern plays in which composers will discover musical possibilities.
III