Chapter I. What Was Bel Canto, Really?

Few terms in the singer’s vocabulary evoke such a bewildering array of meanings as does the Italian phrase bel canto. Depending on the context, bel canto can refer to a style of interpretation or ornamentation, any number of methods of vocal production, a particular Italian musical tradition, a specific opera repertoire, or a certain type of virtuoso or lyric singer. It is used by some to describe artists and techniques of the past, but is often indiscriminately applied by others to those of the present. Philip A. Duey quips that bel canto is “that magic system which every self-respecting teacher of singing professes to teach and which every self-respecting newspaper critic says is an extinct art” (Duey 1951, vi).[1] David Ffrangçon-Davies fears that “the term bel canto is in some quarters perilously near hocus pocus; it is often used as though it were some preparation which singers apply to their voices—as ladies use unguents for their faces—to soften them!” (Ffrangçon-Davies 1968, 21). Blanche Marchesi observes that “people speak of bel canto as they speak of the craters in the moon. Although unaware of what it really means, they think that bel canto is a perfect Italian method” (Marchesi 1932, 154). However, for the purposes of this study, the term bel canto will refer not to a “perfect Italian method,” but rather to the schools of vocal pedagogy that preceded the scientific era, flourishing in Italy and inspiring those of other European nations until the mid-nineteenth century.

Unfortunately, the exact practices employed by teachers of this period are difficult to determine with certainty. Although the bel canto instructors left a record of their school in the form of vocal treatises, much of this literature is far more focused on performance practice issues than on specific teaching techniques. And while there are many written accounts extolling the vocal prowess and virtuosic feats of singers of this era, there are far fewer descriptions of the precise methods used to develop these abilities.

A strong oral tradition, rather than written records, presumably preserved and transmitted much of this crucial knowledge. Manuel Garcia II, in the Preface to the first edition of his Traité Complet sur l’Art du Chant (1841), laments the difficulty of reconstructing actual bel canto practices from the literature available:

It would be curious…to study in detail the teaching practiced in the seventeenth and eighteenth centuries in the schools…of Fedi, Pistocchi, Porpora, Bernacchi…. Unfortunately, that epoch has left us only some vague and incomplete documents of its traditions. The works of Tosi, Mancini, Herbst, Agricola, etc. give us only an approximate and confused idea of the methods then followed (Garcia 1984, xvii).

In 1931, William Earl Brown, the student of another famous bel canto heir, Giovanni Battista Lamperti, also acknowledged this difficulty, concurring that “no definite system of bel canto has descended to us, except advice by word of mouth, from singer to singer” (Lamperti 1931, iii). The vocal pedagogy author William Shakespeare, also a pupil of Lamperti’s, further opines that the bel canto teachers’ failure to publish may have been in some part deliberate, to guard “trade secrets”:

Let it be remembered that it was not the fashion in olden times to publish handbooks on art. The secrets were well guarded, and when the master printed a book it was generally supposed that the student possessed already an elementary knowledge of the subject (Shakespeare 1924, xiii).

Edgar Herbert-Caesari likewise alludes to the bel canto teachers’ possible proprietary attitudes regarding their methods:

Nobody will deny that the glorious traditions of the old school have been dispersed, that the sublime artificers could not, or would not, set down in writing, preferring perhaps jealously to guard the fruit of their vast experience, merely handing it down from generation to generation through the medium of the finished artiste . . . (Herbert-Caesari 1936, xix).

While these writers are correct to point out the scarcity of explicit information about what modern pedagogues might consider “teaching practices,” they neglect to acknowledge the many important basic principles of bel canto technique and teaching practice that are clearly preserved. In fact, there is an almost universal agreement of opinion concerning these foundations in the literature. Duey remarks that certain phrases appear repeatedly in treatises of all nationalities, often word for word:

There was general agreement on the broad precepts of singing, and more important, on the means by which these ends were to be reached. It is really notable to find the identical phrase or sentence turning up in book after book, even in the different languages (Duey 1951, 154).

Bel canto authors, for example, express consistent views on two subjects: the organization of a young vocalist’s studies, and the ideals of vocal tone quality which were their goal. These teachers generally hold that a gradual progression and sufficiently long period of study are crucial for the potential singer’s success. All insist, moreover, that he or she must master the rudiments of music before actually beginning to study voice. Instruction was to begin at an early age, at which time pupils were taught solfege syllables and sight-reading. Their initial vocal training was thus provided not by a voice teacher, but by a solfege master, who was expected to “make the Scholar hit and sound the Notes perfectly in Tune in Sol-Fa-ing” (Tosi 1923, 19). Only when purity of intonation and superior musicianship had been achieved through these studies was the student allowed to pass to vocalizes (i.e., vocal exercises sung on vowels, rather than solfege syllables). In the classic bel canto treatise Observations on the Florid Song (1723), castrato Pier Francesco Tosi warns that the student must be kept on solfege as long as necessary to acquire these two skills, “for if [the Master] lets him sing upon the Vowels too soon, he knows not how to instruct” (Tosi 1923, 29).

The course of study deemed necessary for the complete training of a singer lasted approximately six years. The bel canto teachers deeply believed that this ample time period was essential for the development of sound technique. For example, Tosi and many of his colleagues attribute the decline of vocal art during their lifetimes in part to the insufficient study periods required of young singers. These young and inadequately trained singers, they claim, were often forced before the public prematurely by their greedy instructors, who were eager to reap the financial rewards of their students’ successes. In 1726, Johannes Hiller concurs that “in learning to sing we must force nothing from nature; only gradually, and with thoughtful and persevering diligence, obtain everything we can from her” (Shakespeare 1938, 78).

However, perhaps the most vivid illustration of this commitment to gradual vocal development over a period of years is the legendary anecdote about the great Neapolitan maestro Porpora and his celebrated pupil, the castrato Caffarelli. The story relates how, for six years, Porpora limited his gifted student to practicing the exercises written on a single sheet of music paper. When Caffarelli finally found the courage, after six years, to ask if he might be allowed to sing an aria, Porpora replied that he could choose any aria he liked; he was the best singer in the world. Clearly, the old Italian instructors regarded a slow process of evolution as the safest and ultimately most successful method of vocal development.

The teaching of the bel canto instructors was guided by ideals of vocal tone quality as well as notions of slow progression. The two vocal qualities most sought after were smoothly blended registration and vowel purity. Only two vocal registers were generally acknowledged at this time: the voce di petto (chest voice), also called voce di piena (full voice), and the voce di testa (head voice), or falsetto. (Most bel canto teachers, with the exception of Tosi’s German translator, Johann Friedrich Agricola, did not distinguish between the “head voice” and the “falsetto” or “middle voice” as many teachers do today.[2]) These two registers were to be developed independently, and then carefully blended, so that the “break” between them (located at E or F above middle C in both men and women) ceased to be audible.

In 1558, Giulio Caccini affirms this priority, stating that “the first and most important foundation is how to start the voice in every register…that thereby the quality of the tone be preserved” (Shakespeare 1924, 73). Tosi concurs:

A diligent Master, knowing that a Soprano, without the Falsetto, is constrained to sing within the narrow Compass of a few Notes, ought not only to endeavor to help him to it, but also to leave no Means untried, so to unite the feigned and the natural Voice, that they may not be distinguished (Tosi 1923, 23).

In another landmark vocal treatise of the period, Practical Reflections on the Figurative Art of Singing (ca. 1774), Giambattista Mancini too claims that “the great art of the singer consists in acquiring the ability to render imperceptible to the ear the passing from the one register to the other” (Mancini 1977, 59).

Although these and many other bel canto teachers repeatedly insist on the importance of uniting the registers, exactly how they achieved this is less clear. For example, while Tosi sets the “diligent Master” the challenging goal of uniting his Scholars’ registers, he leaves the details of effecting this union to individual instructors. Whatever the methods he recommended, apparently they were not in common use at the time:

If all those, who teach the first Rudiments, knew how to…unite the feigned to the natural Voice, there would not be now so great a scarcity of Soprano’s [sic] (Tosi 1923, 26-27).

Mancini’s best advice on how to remedy imbalanced registration is “to follow the natural instinct, but never to force nature” (Mancini 1977, 60). He also suggests that when one register is stronger than the other, the student should simply sing more quietly in that register until the lagging one had been strengthened:

. . . the most certain method to help unite them is for the scholar, without losing time, to undertake to establish in his daily studies the manner of holding back the chest voice [the stronger register] and of strengthening little by little the unfriendly notes of the head [the weaker register], in order to render the latter equal to the former in the best possible way (Mancini 1977, 39-40).

However, John Micksch, a student of the noted Italian tutor Bernacchi, does hint that “the joining of the registers can only be attained through the repose of the mouth, tongue, and throat whilst singing,” claiming that “the slightest movement of these three organs disturbs the imperceptible joining of the registers” (Shakespeare 1924, 84).

Another probable means of encouraging blended registration was the practice of the messa di voce.[3] Many authors emphasize the need to perfect this skill on every note throughout the entire compass of the voice. They suggest that the singer pay special attention to those tones which can be rendered in both registers. On these tones, the singer must pass from falsetto to chest voice and back to falsetto again as the dynamic level requires. Mancini instructs his students in the proper mouth opening that will facilitate the various dynamic levels of this exercise:

Do not doubt that in the beginning the scholar will find no little difficulty in the execution of the swell and diminish with equal gradation. But this difficulty will be in part reduced, if in doing the exercises he will fix the mouth well. . . . The mouth should be scarcely open when starting the note, which helps the voice very much in coming forth sweetly, and then gradually it should be reinforced by opening the mouth until it reaches the limits prescribed by art (Mancini 1977, 45).

However, beyond these seemingly simple instructions, modern readers can only assume that the task of uniting the registers was accomplished through repeated practice—in other words, by trial and error.

Beginning in the most elementary stages of voice training, the bel canto instructors insisted on vowel purity, as well as on balanced registration, for proper voice production. Even bel canto predecessors such as Christoph Praetorius, studied exhaustively by Bernhard Ulrich in his book Concerning the Principles of Voice Training during the A Cappella Period and Until the Beginning of Opera (1474-1640) (1910), stressed the importance of “pure” vowels. Tosi is most adamant on this point, instructing the master to be certain “his pupil’s tones, when singing solfeggi, are produced purely” (Tosi 1923, 25) and exhorting the student to “pronounce the vowels distinctly, or he has not got out of the first lesson” (Reid 1950, 82). Mancini also emphasizes the necessity for pure vowel sounds:

The student must not forget to plant the vowel firmly upon the tone on which he is to execute, in order that the beauty of the tone not be impaired (Mancini 1977, 163).

Cornelius Reid, in his modern day review of Bel Canto: Principles and Practices, attributes similar attitudes towards vowel purity in this era to Caccini, Doni, Herbst, Bovicelli, and Zacconi as well (Reid 1950, 39).

It is equally clear that some vowels were considered more conducive to beautiful tones than others. Ulrich notes voice teachers’ preference for open a, e, and o vowels as early as 1613 in the works of pedagogues such as Diruta and Cerone, and reports that Roman and Neapolitan coloratura singers completely eliminated the closed e and u vowels from their exercises because they were “difficult and ugly in their pronunciation” (Ulrich 1973, 134). Tosi encourages his students to study on the three open vowels, the open a in particular (Tosi 1923, 29). Mancini speaks of the “forbidden vowels” of his profession, claiming that the closed e and u should be practiced only “because sometimes one must sing on them” (Mancini 1977, 163). Like Tosi, he suggests that most practice should be restricted to the open a.

From these sources, it is evident that the bel canto instructors had learned from experience that carefully selecting or adjusting a student’s vowel sound would improve his tone. In other words, the bel canto teachers believed that a well-produced tone was inextricably linked to a correct vowel sound. The student, then, through the instructor’s verbal suggestion, or perhaps by following his vocal example, would achieve the correct vowel and practice it over and over, until it became ingrained. Thus, by training the ear to be sensitive to registration and vowel considerations, the bel canto teachers established their students’ control of tone quality.

In addition to their agreement as to the goals of balanced registration and vowel purity, bel canto pedagogues were also of the same mind regarding the vocal faults that could mar this purity. In particular guttural, throaty production and nasal production were equally deplored. As early as 1474, Conrad von Zabern naively reflects that “singing through the nose is not very beautiful”:

Since among the various body organs which work together to produce the tone, the nostrils are never mentioned, it is a sign of inadequate training if one is not satisfied with the mouth and other natural tools, but sends the voice through the nose (Ulrich 1973, 43).

In 1668, Bénigne de Bacilly warns of the dangers of studying with a “teacher who sings through his nose and performs with his tongue, since these defects are easily communicated” (Bacilly 1964, 31). Bovicelli and Praetorius also condemn those who “send the voice through the throat” or “keep the voice in the throat” (Ulrich 1973, 43). Tosi instructs singing masters to:

. . . attend with great Care to the Voice of the Scholar, which . . . should always come for the neat and clear, without passing thro’ the Nose, or being choaked [sic] in the throat; which are two of the most horrible Defects in a Singer, and past all Remedy if once grown into a Habit (Tosi 1923, 22).

Mancini, naming the three most “monstrous effects” of incorrect singing, lists the guttural and nasal timbres as the top two offenders:

First, instead of putting the voice into the air, they carry it into the throat: second, it makes the youth sing viciously in the nose; and for the third, it makes the pronunciation lisping and stammering (Mancini 1977, 29-30).

And finally, Richard MacKenzie Bacon (1776-1844), an early nineteenth-century British heir to Italian bel canto traditions, writes that “if the throat be kept in the slightest degree too much extended or too much closed, the tone will be guttural—a certain action of the nose in producing it will make it nasal.” He adds that these faults, if not corrected in the student at the first sign of error, become ingrained and “can be removed, if at all, only by labour indescribable” (Bacon 1966, 90).

Bénigne de Bacilly’s dire warning to avoid teachers who sing with the defects of throaty or nasal timbre provides an insight into one of the few specific pedagogical methods that actually is often recommended by writers from the sixteenth century until Manuel Garcia II: imitation. Ulrich claims that during the a cappella period (1474-1640) “teaching methods of singing were the same as those of the other free arts: imitation and practice,” and suggests that “the teacher himself was the best example to imitate” (Ulrich 1973, 5). He notes that this view was made clear in works by Coclicus (1552), Cerone (1613), and Praetorius (1614-1620). In the second volume of his work Harmonie Universelle, Mersenne (1637) gives specific instructions to the teacher on how to use imitation during lessons:

First let the teacher sound for the pupil a clearly audible tone, best of all with his voice. . . . Let the student imitate the one, then another one, always using the same syllable, e.g., “la”; then let him accustom himself to imitate every tone he hears. The teacher should then point out the audible difference between a good and bad imitation of the tone as well as the difference between a firm and even tone and an uneven, wavering one (Ulrich 1973, 8).

Bel canto masters evidently continued this practice of their predecessors. Lucie Manen, in her review of Bel Canto: The Teaching of the Classical Italian Song-Schools (1987), claims that this method belonged to Giulio Caccini, the Florentine composer who was also one of the most sought-after voice teachers in that city:

Like all singing teachers of his time, Caccini seems to have taught in the manner also adopted by the later masters of the classical Italian schools of the 18th and 19th centuries. They first of all gave a demonstration, then listened with a finely tuned ear while the pupil imitated it, making continual corrections until the pupil was able to achieve the correct production and quality of timbre (Manen 1987, 18).

Mancini, who taught over a century later than Caccini, defends his use of imitation by enumerating the other renowned singing masters who also employed it:

The easiest method, and one from which I have derived good results in making scholars aware of the proof of the errors they have committed, seems to me to be to imitate (counterfeit) faithfully the defect of the scholar. Thus the scholar can have proof of the error of singing through the nose, of singing in the throat, or with a thick, or crude, or heavy voice. . . . You should not believe that this method of imitation began with me, or with masters of our own day; it was this very method which Angelini Bontempo of Perugia mentions in connection with the celebrated Fedi . . . (Mancini 1977, 32).

Bacon also writes of the benefits of imitation in teaching more difficult passages:

As we penetrate into the mazes of execution it becomes more difficult to give any rules in words for the direction of practice. It is a very extraordinary fact, that an individual will catch in a moment, from hearing a thing done, that which he never, by the force of his own genius, could have been able to attain (Bacon 1966, 97).

The teacher’s ability to demonstrate correct vocal production was evidently one of the primary requirements for successful instruction.

Even with insights such as the use of imitation, it is admittedly difficult to reconstruct the breadth of bel canto teaching methods with any assurance. However, based on the surviving vocal literature of the era, one can state with relative certainty that these practices did not hinge upon physiological knowledge of the internal vocal organs, nor were they designed to teach a mechanical control of them. For though by the eighteenth century the dissection of cadavers had allowed scientists to document the anatomy of the vocal organs fairly well, neither scientists nor voice teachers correctly understood their physiology, as no means of observing the internal vocal organs in action had yet been devised. Thus, theories of phonation, when offered at all, were most often superficial and based on incomplete or incorrect information. Others merely repeated the speculation of ancient physiologists.

The greatest of these physiologists was Galen, who is considered the father of laryngology. In the second century A.D., Galen advanced the idea that the pitch and volume of vocal sounds was dependent upon the width of space between the vocal cords.[4] This incorrect theory was generally regarded as truth for well over a millenium (Duey 1951, 14), even by otherwise great scientific thinkers and anatomists. For example, Leonardo da Vinci compares the throat to an organ pipe, concluding “that the pitch of the human voice is dependent of the length and diameter of the trachea” (Duey 1951, 16). Likewise, the erudite German instructor Agricola, best known for his annotated 1757 translation of Tosi’s Observations, accepts and painstakingly elaborates on Galen’s mistake, explaining how the “opening of the windpipe…can be widened and drawn together so it can thus produce high and low tones” (Duey 1951, 131).

Agricola’s mistake is particularly remarkable, given that over a decade earlier, the French anatomist Antoine Ferrein (1693-1769), professor of anatomy at the Jardin du Roi in Paris, had almost correctly identified the role of vocal cords in determining the pitch of vocal sounds. In the 1741 Mémoires de l’Académie Royale des Sciences, Ferrein was able to show

. . . first, that in order to phonate, the lips of the glottis had to come together, second, that vibration of the lips was the essential factor since by touching them the sound stopped, and third, that difference in tension of the edges of the glottis caused the changes in pitch (Duey 1951, 17).

However, Agricola was not the only pedagogue writing after 1741 who was either unaware of Ferrein’s findings or who dismissed them. For example, in the 1770s, Mancini simply observes that the muscles of the larynx contract strongly on high notes, while relaxing on lower ones. Naming the components of the voice as the larynx, the glottis, the uvula, the palatal veil, the tongue, the teeth and the lips, Mancini explains that

In speaking these organs are quiet and natural, but in the action of singing they are held to constant toil, and the most fatigue is in the muscles of the larynx: these direct the voice, condensing to produce the high notes and dilating for the low notes (Mancini 1977, 17).

Mancini goes on to incorrectly attribute the strength of the voice to the size of the lungs, the “broadness” of the trachea, and the width of the larynx. Even as late as 1823, Isaac Nathan, a student of the famous Neapolitan master Porpora, also correlates pitch changes in the voice with the supposedly variable lengths of the human trachea as well as to the size of the glottal opening:

The varieties of tones in human voices arise partly from the dimensions of the trachea or windpipe, which, like the flute, the longer and narrower it is, the more acute is the sound it gives, but principally from the head of the larynx, or knot of the throat, called pomum adami, the tone of the voice being more or less grave as the rima or clef therein is more or less open . . . (Nathan 1968, 88).

Other theories regarding aspects of phonation are even more mistaken, illustrating just how far off the scientific mark many otherwise noted vocal pedagogues were in their understanding of the physiology of the voice. For example, Mersenne, in his 1637 publication of Harmonie Universelle, ascribes a singer’s facility with coloratura to the elasticity of his glottis and the activity of surrounding blood vessels:

Those who have a more flexible glottis (the opening of which opens and closes more flexibly) are adapted for passages and trills. Those, however, with a hard and dry glottis cannot perform coloratura. . . . Also the activity of the blood vessels that feed the vocal organ must be very strong. In this sense one can say that those who have the mastery of coloratura singing have an elastic glottis because their glottis opens and closes more easily than those of other people (Ulrich 1973, 147).

Carré, in his Nouvelle Méthode, published in 1719, actually attributes the beauty of vocal tone to the size of the singer’s uvula:

It is certain that the little cartilagenous membrane resting on the windpipe . . . when it is correctly proportioned (this is to say when it is neither too large and heavy, nor too small) contributes greatly to the strength and beauty of the voice (Duey 1951, 134).

Isaac Nathan concurs with and elaborates on a similarly fantastic “uvula theory”:

. . . on the flexibility of the uvula, and the muscles connected with it, depend both the perfectibility of the shake and execution. . . . The uvula, a peculiarity confined exclusively to singers, is singularly mobile; its apex being also finely pointed—whilst in those persons who are not famed for vocal talent, this organ is found to be possessed of scarcely any muscular power, and its apex, or lower part, round and obtuse. It is strange that this never-varying rule should have escaped the observation of men, who have devoted their lives to physiological research (Nathan 1968, 261).

He further erroneously claims that while the falsetto voice (i.e., the middle voice) is governed by the “contracted aperture of the mouth,” the feigned voice is produced at the back part of the head and throat, “where the uvula is situated” (Nathan 1968, 144-5).

The inclusion of such supposedly scientific speculations in bel canto works was evidently motivated by curiosity, and not a desire to find information that would aid in the singer’s training. Duey summarizes that “any emphasis on the physiological study of phonation in the art of singing before 1800 was based on false or incomplete theories,” and concludes that, therefore, any “successful methods of teaching used were, perforce, entirely empirical in nature” (Duey 1951, 18).

However, unlike the mechanics of the singer’s internal organs of phonation, those of the exterior organs (mouth, lips, teeth, tongue) were better understood and of great concern to the bel canto teachers. In particular, the proper position of the mouth during singing was of special interest, and on this topic all the major authors concur. Tosi, for example, directs that “the mouth . . . ought to be composed in a manner . . . rather inclined to a Smile” (Tosi 1923, 26). Mancini maintains that “every pupil must shape his mouth for singing, just as he shapes it when he smiles” (Mancini 1977, 93). And Manfredini submits that “the correct way is to keep [the mouth] open as in the act of smiling” (Duey 1951, 105).

The bel canto masters also consistently expressed their conviction that any facial grimace or unnatural quirk of the mouth must be strenuously avoided, as such habits impede the “natural and spontaneous” action of the vocal organs (Mancini 1977, 96). Likewise, Bacon cautions students to “guard with the severest attention the first appearance of any mal-formation [sic] of the mouth or distortion of any kind; for although Billington elevated one side of her mouth on high notes, and Catalani has a constant quivering of the under jaw during her execution of rapid divisions, these are deformities which might have been avoided if corrected at first” (Bacon 1966, 102). Tosi suggests that students use a mirror when practicing in order to ensure that such distortions do not become ingrained:

When he studies his Lesson at Home, let him sometimes sing before a Looking-glass, not to be enamoured with his own Person, but to avoid those convulsive Motions of the Body, or of the Face (for so I call the Grimaces of an affected Singer) which, when once they have took Footing, never leave him (Tosi 1923, 89).

Bel canto writers also had some understanding that unnatural or unnecessary tensions of the tongue during phonation would negatively affect the resulting vocal tone. As early as the late sixteenth century, pedagogues were instructing students in the proper disposition of the tongue while singing. Maffei (1562) counseled students to “stretch the tongue out until the tip of the tongue touches the base of the lower front teeth” (Ulrich 1973, 55). Coclicus (1552), a disciple of Josquin des Prez, wrote that

. . . the student must work by the sweat of his brow and force himself to repeat his exercises daily until, through knowledge and practice, he does not once move the tongue, but sings with the correct attack. . . . It denies a false position such as the raising of the back of the tongue and any movement that hinders the tone stream (Ulrich 1973, 56).

Agricola (1757) also describes obstructions in the throat that can be caused by the tongue, “if one retracts the tongue when it is not necessary, or bends it, when it should really be in the mouth flat and straight” (Duey 1951, 106). Mancini deplores the “monstrous position” of singers who “put the tongue on an equal basis with the lips”:

. . . since the tongue is not lying in its accustomed place, the emission of the voice cannot be sonorous, because it strikes the palate and remains choked in the throat. . . . [When] the emission of the voice is impeded by the enlargement of the tongue, instead of the defect of the throat, it acquires the vice of the nose . . . (Mancini 1977, 29-30).

Mancini therefore counsels students to make sure their tongues instead “rest quietly” while singing, perhaps even with a slight groove down the center (Mancini 1977, 129).

As has been noted, accurate information about the anatomy of the internal vocal organs such as the larynx and pharynx was scarce, and speculation regarding their physiology was usually far from scientific truth. However, it is quite clear that vocal pedagogues judged even the available information about the internal vocal organs (correct or not) to be relatively insignificant for their teaching. Some actually state flatly that the inclusion of such detail in a vocal method would be inappropriate. Mengozzi, for example, at the outset of his Méthode de Chant (1803), declares that “a scientific definition of [the larynx] would not enter usefully into the organization of a singing method” (Duey 1951, 137). A decade earlier, in 1791, Kempelen compares the uselessness of a voice teacher including a complete physiological study in a vocal method to a violin teacher who began a lesson by giving the student a description of all the parts of that instrument, how they are connected, the kinds of wood used in each, the construction of the sounding boards, strings, bridge, pegs, etc., as well as the muscles and tendons used in fingering and all the other complicated mechanics of violin-playing (Duey 1951, 136).

One further indicator of how the bel canto pedagogues viewed anatomical and physiological information about the internal vocal organs is the fact that such knowledge is not included as a qualification for singing masters in virtually any of the dozens of lists of such qualifications included in the literature. For example, among Bacilly’s carefully articulated six prerequisites of good voice teachers in his chapter on “The Choice of a Voice Teacher, and the Qualities He Ought to Have,” none has anything to do with knowledge of anatomy or physiology (Bacilly 1964, 31-2). The one notable exception to this disinterest is Tosi’s translator Agricola, who asserts that

The knowledge of the vocal organs is always very useful to the singer, and especially to the teacher, and in many cases indispensable. For even when nature has adorned a singer with the best qualities, the knowledge of physiology is necessary to prevent all damages that might be done through ignorance. But when a teacher finds natural faults and defects in a voice how can he successfully battle with them if he is unacquainted with the seat of the evil? (Browne and Behnke 1890, 5).

However, since Agricola’s own knowledge of the vocal organs was later proven to be partly inaccurate, any ability he had to battle the vocal defects in his students was clearly not due to his physiological knowledge, regardless of what he may have believed. Furthermore, Duey points out that Agricola’s real attitude toward the whole problem of voice production is more clearly revealed in the following statement:

While the natural scientists remain silent, it is not up to us, who do not consider the organs of the voice from the point of view of the anatomist or physician but from the point of view of the singer, to decide whether Mr. Ferrein’s discoveries through his many experiments are correct or not, as little can be said against his conclusions derived therefrom (Duey 1951, 133).

Apparently, even Agricola felt that it was not part of a singing teacher’s job to assess the accuracy of scientific information available, as long as that information led to results with the student.

Finally, those few pedagogues other than Agricola who do mention an understanding of vocal “science” as an important prerequisite for a singer or teacher define this term in a very different way from a modern linguist. Rather than referring to issues of anatomy, physiology, or acoustics, the term “science” is often used instead to refer to a highly developed knowledge of current performance practice, combined with good taste:

The term Science, in its particular acceptation, when applied to singing, appears to be very difficult to define. . . . Everyone talks of “the science” of a singer, but there are few who have ever stopped to consider what they purpose to express. Science appears to me to imply the perfect union of taste and knowledge—the complete combination of style and manner—a thorough acquaintance with the rules of art, and a power of reducing them to just practice (Bacon 1966, 86).

So even bel canto teachers who considered themselves to be “scientists” would likely not be considered so by today’s standards.

In addition to their relative indifference about internal vocal anatomy and physiology, the bel canto teachers further believed that one need not attempt mechanically to control the vocal organs, emphasizing instead that the action of these organs is governed instinctively. For instance, in 1760, Jean Philippe Rameau recognizes that “we cannot do as we please with the larynx, windpipe, and glottis, nor can we see their different motions and changes at each sound we wish to produce,” but stresses that “we know, at least, that it is not necessary to force them into making these changes, that it is necessary to allow them the freedom of following their natural movements” (Duey 1951, 136). Taylor notes that in the Méthode de Chant du Conservatoire de Musique, published in Paris in 1803, the mechanical management of vocal organs was “not even thought of”:

There can be no question that this Méthode represents the most enlightened and advanced thought of the vocal profession of that day . . . To mechanical rules less than one page is devoted. Respiration is the only subject to receive more than a few lines (Taylor 1922, 327).

Duey attributes this attitude to bel canto pedagogues in general:

There were no attempts at the conscious control of particular muscles. The singer was urged to keep a natural and free muscular balance of all his physical faculties. . . . All the external mechanical forces as well as the inward nervous and muscular controls were given a set of relationships that encouraged their free and unhampered functioning (Duey 1951, 154).

The relative absence of any such physical directives in the bel canto literature prompt Monahan to speculate that a lack of awareness of the internal vocal organs was actually a deliberate goal of the pedagogy:

Discussions on the nature of phonation and on relative merits of local control in the throat simply do not appear until [the mid-nineteenth century]. These facts would tend to prove that teachers of the eighteenth and early nineteenth centuries followed the bel canto adage, “An Italian singer has no throat,” patently refusing to draw the pupil’s attention to this area (Monahan 1978, 99).

But perhaps the clearest summary of the bel canto attitude toward including science in voice teaching was articulated by Giovanni Bontempi in his work Historica Musica (1695). In this treatise, Bontempi clearly counts the rational investigations of the physiologist unnecessary to explain or justify the methods of the singing teacher:

The physiologists expound these differences, caused by position, passage, shape, air, expiration, and all the conditions of the larynx; based on the immutable foundation of incontestable reason. Our opinion is, that everything which is derived from experience has no need of reasoning . . . while they go on investigating nature in order to find it out with reason, we, without philosophizing further upon it, shall be content to understand it with the teaching of experience itself (Duey 1951, 127).

Bontempi here highlights the most fundamental bel canto principle of all: faith in the instructive value of practical experience. Clearly, in the absence of reliable scientific information, systems of training could only be developed around personal observations. Teachers learned through experience to identify what conditions were most favorable to the emission of the voice. Mancini’s discussion of proper mouth position, for example, provides a concrete illustration of such an observational process at work. He justifies his extensive treatment of this subject by asserting that “experience proves that the opening of the mouth is what directs and regulates the voice” (Mancini 1977, 92). He then explains that general mechanical rules for the singer are of little merit, given the differing constitution of the vocal organs in different singers:

The rules for the opening of the mouth cannot be general, nor can they be made universally the same for every individual. . . . Differences concerning the organs of the voice compel the teacher to observe diligently in what size of the opening of the mouth the voice comes out the clearest, purest, and fullest [emphasis added]. Thus he determines the size of the opening for the correct position of the mouth (Mancini 1977, 90).

Mancini later reinforces this point by restating it in general terms: “although a method is a good one in itself, it cannot be applied in every case with good results . . . each fault in the voice requires a different remedy to be applied to the origin of the fault” (Reid 1950, 97). In correcting a fault, Mancini cautions the teacher to avoid precise rules in favor of practical demonstrations:

In giving the precise rules to a student let the teacher not only tell him and explain to him, but let him illustrate his meaning by making himself an example . . . Let the experienced teacher follow this method and he will soon be convinced how much more preferable are practical demonstrations to general rules (Duey 1951, 153).

Bacon echoes these same sentiments:

It is therefore a momentous concern to ascertain in what exact position of the organs the tone is produced with the least effort, always having regard to its purity and beauty. I confess I doubt the possibility of determining this by any known formula. It must be a matter of experiment and experience (Bacon 1966, 97).

Clearly, bel canto teachers were highly sensitive to the varying natures of their pupils, and would tailor their teaching methods to suit each one’s individual needs.

Modern scholars who have studied bel canto sources in detail all comment upon the empirical nature of teaching practices during this period. For example, Elster Kay, in his 1963 review subtitled A discursive and technical essay on the traditional Italian physical style of singing, asserts that teachers during this period eschewed predetermined methods in favor of certain solfeggi and arpeggio, “not because their teachers had any idea of the physiological processes involved in singing them, but because it was well known that such methods were effective . . . ” (Kay 1963, 26). Lucie Manen agrees that bel canto pedagogy was not based on “any explicit theoretical method” but rather, on “trial and error”:
Its teachers, the maestri, were themselves expert performers. They instructed their pupils in the same way that they themselves had learned, by trial and error, until their pupils were able to achieve the right vocal quality. This entirely empirical method, with its unremitting process of trial and correction, relied heavily on the accuracy of the pupil’s hearing and the acute ear of the teacher (Manen 1987, 3).

Finally, Duey, the highest authority on bel canto methods, provides one final confirmation of these conclusions:

The vocal pedagogy of the bel canto period was based on empirical methods. There is a wealth of evidence that the singing experience of the teacher was a most important part of the teaching method. . . . The teacher set himself up as an example to be observed and imitated by the pupil . . . (Duey 1951, 152).

In sum, the bel canto school encouraged a gradual process of vocal discovery in which the natural gifts of each singer were eased into awareness. Teaching methods were arrived at on the basis of practical experience, and were specific to individual students. Scientific knowledge of the voice, though admittedly scarce, played no appreciable role in the shaping of teaching methods, and, in fact, was generally viewed as irrelevant to vocal success. Control over the internal vocal organs was obtained indirectly, through a blending of the vocal registers and strict observance of vowel purity, but never through deliberate attempts at mechanistic adjustment.

For hundreds of years, these simple guidelines and goals held sway over the field of vocal pedagogy. While the concepts were born in Italy—the singing capital of the world during this period—they radiated out to the rest of Europe with remarkably little variation. The approach to voice teaching was so unified that, in fact, if “a student in the Neapolitan school left to study at the Roman or Bolognese school, there would be absolutely no change . . . ” (Herbert-Caesari 1936, 6). But it would take the career of only one man, the formidable Manuel Garcia II, to shake the foundations of this venerable tradition, and usher in a new era in vocal pedagogy that was as divided as the prior era had been unified.


[1] Duey’s comprehensive work on Bel Canto in Its Golden Age: A Study of Its Teaching Concepts (1951) contains an extensive history of the term bel canto, tracing its adoption into general usage to the time of Wagner, when it was pitted against the dramatic, declamatory vocalism that the Wagnerian repertoire was believed to require.
[2] Duey notes that the existence of these registers was recognized and discussed during the Middle Ages, even before the true bel canto period, by John of Garland (ca. 1193-1270) and Jerome of Moravia (ca. 1250), both of whom recognized three vocal registers which they associated with the chest, throat, and head (Duey 1951, 33).
[3] The messa di voce is a vocal exercise in which the singer holds a single tone on a single vowel, beginning at pianissimo, gradually making a crescendo to forte, then gradually returning to pianissimo.
[4] The pitch of vocal sounds is actually determined by the speed at which the vocal folds are abducted and adducted. This speed is determined in turn by their length, tension, and segmentation, as well as by the pressure of the singer’s breath which sets them in motion.