PAS-Conference October 3-5, 2002
Home Up

Keynote: Historical Antecedents to Current Issues in Voice Science

James Stark

I feel very honoured to speak here today, since I cannot imagine a more distinguished or informed audience. I was surprised, as perhaps you were, that a musicologist and singer would be invited to give a keynote address at what is largely a voice science conference. But upon reflection, it makes sense that, in a retrospective conference such as this, we should begin at the beginning, that is, with the origins and historical development of what we call the "classically trained" singing voice. As a musicologist and singer, I have spent my whole career studying the historical literature on singing, and seeking a rapprochement between voice history and voice science.

A 1967 article in Folia Phoniatrica opened with this statement: "The art of singing has achieved a high level of refinement and sophistication -- not so the science" (Rubin, LeCover, and Vennard, 29). This statement reflects the fact that the art of classical singing is hundreds of years old, whereas the field of voice science is a far more recent discipline. The objective measurements of classical singing techniques provided by researchers using sophisticated technology have been a revelation after so many centuries of ambiguous and subjective descriptions of singing. Notwithstanding the difficulties involved in interpreting these historical treatises, I nevertheless wish to show you that they have a value that is highly important for singers, and that they can help to define the vocal issues that should concern today’s voice scientists. Taken together, the historical literature on singing and modern research in voice science span a broad range of issues that have much in common. It is this commonality that forms the focus for today’s paper. For as long as there has been "classical singing," that is, operatic and concert singing, dating back to the late sixteenth and early seventeenth centuries, similar vocal issues have kept presenting themselves over and over again in one way or another, and the vocal problems that concerned the earliest opera singers continue to concern today’s singers, teachers, and voice scientists. My presentation today will call attention to just a couple of the most important issues, namely, the interrelated subjects of voice quality and glottal closure.

I will begin by taking a look back at some historical views of voice quality, particularly the matter of "brightness" or "edge" in the trained voice, an issue that continues to concern us today. It is well known that the Renaissance was the period of Franco-Netherlandish polyphony, that is, music where choral blend and the homogeneity of the voice parts formed the sound ideal (Klang-Ideal). But in the later sixteenth century, Italy imported some important composers from the Low Countries, and the Italians quickly adapted Franco-Netherlandish polyphony to their own ends, which meant that they began a process of conversion from a cappella part-music to accompanied solo singing with virtuosic qualities. With this conversion came the earliest operas and accompanied art songs. Whereas the choral singer sought blend, the solo singer sought strength, virtuosity, and individuality. The world of voice now appeared to be split between two ideals of singing, namely, the older choral ideal and the newer soloistic ideal.

A key figure who observed this split was the Italian Lodovico Zacconi (1555-1627), who was a singer and a maestro di cappella in Venice, as well as a composer, author, priest, and inventor -- a true Renaissance man. In his Prattica di musica of 1592, Zacconi included a chapter titled "On the types of voices that should be selected in order to make good music." He differentiated between voices that were "dull" (obtuse, mute), and those that had a certain "bite" or "sting" (mordente). He was speaking primarily of male singers, both falsettists and those who sang in the chest register. He expressed his preference for the bright and ringing chest voice, while discrediting both dull and shrill voices. He wrote, "Among all voices, one must always choose ... the chest voices, and particularly those with the above-mentioned delightful biting quality that pierces a little, but does not offend; and one must leave aside the dull voices and those that are simply falsetto, because the dull voices cannot be heard among the others, and the falsetto is overbearing" (Zacconi 1592, fol. 77).

In 1602, Zacconi’s remarks were echoed by Giulio Caccini (1546-1618), the most famous singer and voice teacher of his time. In his Le nuove musiche Caccini wrote a lengthy preface touching on many matters of vocal technique, musical style, and vocal expressivity. Regarding voice quality, he said that one should sing with a "full and natural voice" (voce piena e naturale), avoiding falsetto (la voce finta), ... without being constrained to accommodate himself to others" (Caccini 1602, viii). This last remark was clearly a reference to the fact that the choral singer must necessarily accommodate himself to others in order to achieve choral blend, whereas the solo singer sought individuality. Numerous later authors followed Zacconi and Caccini in their preference for a full, bright voice quality, well suited to the large theatres and virtuosic demands of operatic singing (Stark 1999, 194-197). Caccini and his large number of celebrated pupils and disciples played a major role in establishing what is often referred to as "the old Italian school of singing" (Stark 1999, 190-197).

At this point I would like to demonstrate my interpretation of the dull voice without mordente, as well as the full natural voice with mordente. Power spectra and long time average spectra will appear on the screen, and will show the increased acoustic energy in high-frequency components in the bright tone quality of the chest voice.

I will demonstrate using two vowels, [a] and [u].

Play A (Dull) Play A (Bright) Play U (Dull) Play U (Bright)

 The description of good voice quality took a step forward with the concept of chiaroscuro during the eighteenth and nineteenth centuries. Chiaro referred to the bright edge in the trained voice, while scuro referred to a simultaneous dark and round quality in the tone. The word was borrowed from painting, where the conflict between bright and dark was often striking and dramatic (we need think only of Rembrandt). Every sung tone was supposed to have a bright edge as well as a dark and round quality in a complex texture of vocal resonances. The term was used as early as 1774 in an influential vocal treatise titled Pensieri e riflessioni pratiche sopra il canto figurato (Practical Reflections on Figured Singing) by Giambattista Mancini (Mancini/Foreman 1967, 42). This manual for singers went through several editions and was translated in whole or in part into French, German, and English (Stark 1999, 33). In it, Mancini gave instructions in how to practise slow scales, saying "this exercise will make [the singer] master of coloring at will any passage with that true expression which forms the cantilena colored with chiaroscuro, so necessary in every style of singing" (Mancini/Foreman 1967, 33). Chiaroscuro remained the vocal ideal for Giovanni Battista Lamperti (1830-1910), one of the most famous singing teachers of the late nineteenth century. He was the son of the equally celebrated Francesco Lamperti (1813-1892) who wrote several voice manuals and taught many famous singers (Stark 1999, 42-43). Giovanni Battista said, "Although you may acquire a wide range of voice, you cannot modulate the sounds until the resonance of your tone becomes round and rich, chiaroscuro ... The ‘dark-light’ tone should be always present" (Brown, 1957, 38-39).

A more recent description of chiaroscuro was offered by Richard Miller, who said:

An extensive terminology exists, in several languages, for the description of variations of vocal timbre found within the several [national] schools. One such term is chiaroscuro, which literally means the bright/dark tone, and which designates that basic timbre of the singing voice in which the laryngeal source and the resonating system appear to interact in such a way as to present a spectrum of harmonics perceived by the conditioned listener as that balanced vocal quality to be desired--the quality the singer calls "resonant" (Transcripts 1983, 2: 135).

I believe that chiaroscuro is a voice quality so distinctive that not only conditioned listeners, but casual listeners as well will quickly associate it with opera and concert singing. It can hardly be confused with vernacular styles of singing or with the choral voice. Chiaroscuro is a voice quality that bears within itself a dynamic that is both complex and compelling.

At this point I would like to demonstrate my characterization of chiaroscuro. Once again, I will sing a simple descending five-note scale. The first time I will sing with a bright tone -- too bright to be acceptable as a desirable voice quality. Watch the screen for the energy in the high-frequency components of the tone. The second time I sing the pattern, I will use a dark, sepulchral tone which is also unacceptable as a desirable voice quality, and which is relatively lacking in high-frequency components. It is truly obtuse. The third time (with any luck!), I will sing with a voice quality that combines the bright and the dark qualities of the first two examples. Listen carefully for this mixture of resonances, which is the result of the interaction between the glottal source and the vocal tract. This, I believe, is chiaroscuro. Once again I will demonstrate with two vowels [i] and [u].

 It was the great nineteenth-century voice teacher Manuel Garcia II (1805-1906) who first attempted to explain voice quality in a systematic way by making a clear distinction between the glottal source and the resonances of the vocal tract. In 1841 he established this principle in Part One of his celebrated Traité complet de l’art du chant, where he wrote: "Each modification introduced into the mode of producing the vibrations engenders a different timbre, and each modification which the tube of transmission undergoes modifies the original timbre" (Garcia 1847, I: 16; Garcia/Paschke 1984, 29; the italics are Garcia’s). He reiterated this in Part Two of his Traité when he said, "The timbre depends not only upon the modifications which the tube imparts to the sonorous waves, but also upon the place where these vibrations are born; that is to say that the voice receives its first character in the glottis, then from the numerous modifications of the pharyngeal and buccal cavities" (Garcia 1847, II, 54; Garcia/Paschke 1975, 153). His most complete statement is found in his Hints on Singing of 1894:

The student should thoroughly understand that the ring or dulness [sic] of sound is, in effect and mechanism, completely distinct from the open and closed timbres. The ringing and dulness are produced in the interior of the larynx, independently of the position, high or low, of this organ, while the open or closed qualities of the voice require the bodily movement of the larynx, and of its antagonist the soft palate. Hence, any timbre may be bright or dull. This observation is most important for the expressive qualities of the voice" (Garcia 1894, 12).

Garcia referred to the bright glottal signal as voix éclatante, and the dull or "veiled" sound as voix sourde. The sounds as modified by the vocal tract he referred to as either clair, which referred to a high laryngeal position, or sombre, which required a lowering of the larynx and an expansion of the pharynx (Garcia 1847, 15; 1894, 11). It is clear that Garcia provided an early description of what would later be called the source/filter theory of phonation. Garcia’s vocal method was intended to teach singers how to couple the voice source with the resonance tube so that "it puts the singer into possession of all the ‘tints’ of the voice, and indeed initiates him into all the secrets of voice production" (Garcia 1894, 7).

This last point is a crucial one. Up to now we have been dealing with authors who had a clear preference for a full, bright tone quality that was balanced by a darker resonance. Some of the authors that I have cited have implied that there were only two qualities of voice: dull and bright, an oversimplification that had begun with Zacconi. The crux of Garcia’s method, however, is found in the numerous ways the glottal source and the vocal tract interact in creating the "tints" of the voice -- not just two qualities, but many qualities. Garcia devoted considerable space in his treatises to describing how to do this, and he related the various colours to the expressive demands of specific pieces of music. To put this another way, he advocated a particular voice quality that should be used most of the time, but he also described a wide palette of vocal colours that would suit the musical and dramatic requirements of various kinds of music. While Garcia did not employ the word chiaroscuro, he considered the "purest tone" to be that which was emitted with both éclat and rondeur. This, of course, is the equivalent of chiaroscuro.

For Garcia, the best means of achieving a bright, ringing sound was to sing with a shortened glottis. Garcia’s description of the vocal folds makes a clear distinction between the anterior three-fifths of the vocal folds, which are made up of membranous tissue, and the posterior two-fifths, or cartilagenous glottis, which extends from the point where the vocal processes meet to the base of the muscular processes. In 1894 Garcia wrote, "Let us notice henceforth that [the vocal folds] are not homogeneous throughout their length; the posterior two fifths are formed by a cartilagenous prolongation, and the anterior three fifths by a tendon [ligament]." (Garcia 1894, 205). He noted that when only the anterior portion of the folds participated in the vibrations, "the voice takes on a very pronounced brilliance" (Garcia 1847, II, 54; Paschke 1975, 152-3). I will return to this point later.

Garcia’s technique for achieving a shortened glottis was his famous and widely misunderstood coup de la glotte, or "stroke of the glottis," in which the singer closed the glottis firmly just before the onset of phonation, and then allowed only the anterior three-fifths of the vocal folds to participate in the ensuing phonation. Garcia formulated this theory before his invention of the laryngoscope in 1855, but once he was able to view his own vocal folds he confirmed his earlier statements (Garcia 1894, 7).

In many respects, Garcia’s theory of resonance was simple, correct, and sufficient for his pedagogical purposes, but by today’s standards it is incomplete. A fuller explanation of the particulars of vocal resonance had to wait for the development of modern acoustical theory. The foundation of this theory was laid by Hermann Helmholtz in his monumental book, Die Lehre von den Tonempfindung, first published in 1862 and later translated by Alexander J. Ellis as On the Sensations of Tone (2nd ed. 1885). Helmholtz explained that tones consist of a fundamental frequency as well as partials. Regarding the singing voice he made some acute observations. He recognized that a strong voice requires "perfect tightness" of the glottis. He noted that in the male voice there is an energy peak that corresponds to the highest octave of the piano, which he described as sounding like "a clear tinkling of little bells." (Incidentally, F7 sharp, in the top octave of the piano, has a frequency of 2960Hz, the same region as the "singer’s formant.") Helmholtz said, "This kind of tinkling is peculiar to human voices; orchestral instruments do not produce it in the same way either so sensibly or so powerfully" (Helmholtz/Ellis 1885, 116, 103). These remarks clearly foreshadow the more thorough treatment of the "singer’s formant" by Johan Sundberg in his celebrated 1977 article in Scientific American. Helmholtz also discussed at length the many ways in which the vocal tract could alter tone quality and vowel colour. Like Garcia, Helmholtz observed that vowel modification must take place on high female pitches, especially in soprano voices above F5 (698Hz). Further, Helmholtz recognized the importance of high-frequency components in the voice when he said, "Ringing and keen, applied to a quality of tone, imply many and powerful upper partials ...." He also recognized the need for a singer to "occasionally interpose among his bright and rich tones others of a duller quality as a contrast," depending upon the mood and expression of the music" (Helmholtz/Ellis 1885, 103-115). In short, Helmholtz, like Garcia, advocated strong glottal closure and a bright tone quality, and, like Garcia, his observations foreshadowed issues that are important in today’s voice science.

An understanding of formant tuning as an important aspect of voice quality in singing only developed gradually. Wilmer T. Bartholomew (1934; 1945), Fritz Winckel (1953), and William Vennard (1967, 119-120) all maintained that the ideal singing quality has both a low and a high formant, a concept consistent with chiaroscuro. Sundberg and others have gone much further in refining our understanding of formant tuning, especially regarding the singer’s formant and the high notes in the soprano voice. Sundberg wrote:

A long-time-average spectrum of the human voice normally exhibits at least one or more pronounced peaks. [The first] appears around 500Hz and is related to the first formant frequency ... In singing of both sexes, a second peak also appears in the vicinity of 2 or 3 kHz; it corresponds to the singer’s formant (Sundberg 1987, 130).

In the male voice, the singer’s formant is a clustering of the third and fourth, and sometimes the fifth formant. For soprano high notes from F5 upwards on open vowels, the singer tunes the first formant to the level of the fundamental frequency (that is, the first harmonic (F1 to H1) by varying the vocal tract. In many cases this adjustment seems to be made by the singer intuitively, or by just having a good ear. All this is evidence that the understanding of vocal acoustics as explained by modern voice science is one of the finer achievements of the last quarter century.

In the male voice there is another resonance strategy that has received little attention in the voice science literature, but that may be as important as the singer’s formant. This is a type of formant tuning in which the second formant is tuned to the third or fourth harmonic. The bright quality in my own singing owes more to second formant tuning than to the singer’s formant, throughout its range. Spectral analysis carried out by Harm Schutte and D. G. Miller of recordings of famous tenors singing high notes without orchestral accompaniment revealed that on high "covered" notes some tenors used second formant tuning instead of the singer’s formant (Schutte and Miller 1996). The brilliance that comes from a strong second formant has a different character from the singer’s formant, and a trained ear can readily distinguish between them. My own subjective perception is that F-2 tuning is more "visceral" than the singer’s formant. I believe that the issue of second formant tuning has not received the attention it deserves, and that it is important to recognize the importance of this alternative strategy for achieving a bright, ringing tone quality, not only in the tenor voice, but in baritones and basses as well (Miller, D.G., and Schutte, H.K. 1990).

I would like now to return to Garcia’s discussion of glottal closure, not only as it relates to voice quality, but also as it relates to airflow rates and closed quotients, that is, the percentage of time in each vibratory cycle that the glottis is closed. Once Garcia could actually see the vocal folds with his laryngoscope, he observed that when the glottis was "agitated by large and loose vibrations throughout its entire extent," including the cartilagenous portion, the tone emitted was "veiled and feeble." On the other hand, when the firm closure of the arytenoids brought the vocal processes into deep contact, a full and vibrant sound was created. Said Garcia, "This deep contact, which continues even after the vocal processes no longer take part in the vibrations, gives a deep tension to the membranes, increases the depth of their contact, and, as a necessary consequence, augments the resistance they present to the air" (Garcia 1855, 408). He reiterated this when he said, "it is necessary to conclude that the brilliance of the voice results from the firm closure of the glottis after each pulsation ... the glottis should close hermetically after each pulsation; for, as we have just seen, if the air found a constant outlet, then the largest excursions of the glottis and the strongest expenditure of air would produce the weakest tones" (Garcia/Paschke 1984, 27). Garcia thus noted that the bright sound, using firm phonation, used less breath than the dull sound which employed loose phonation. He wrote, "The waste of air can be verified by placing a lighted match before the mouth. The brighter sound does not stir the flame, the veiled one will" (Garcia 1894, 7). Although Garcia’s technology included only a simple angled mirror and a matchstick, his theories were far ahead of his time.

Garcia’s frequent reference to the shortened glottis, to the "pinching" of the glottis, to breath resistance by the glottis, and to the coup de la glotte, in which the glottis was firmly squeezed before the onset of phonation, was unique. The éclat, that is, the high-frequency components in the voice, comprised the "edge" to the tone. To the best of my knowledge, no other vocal treatises up to Garcia’s time had discussed vocal onset and glottal settings in any detail, nor had any authors placed such importance on the method of beginning a tone. The coup de la glotte was the very basis for Garcia’s method, because it had such a profound effect on glottal closure and the ensuing voice quality. Garcia said, "One should insist on this first lesson, which is the basis of the teaching" (Garcia 1847, I:26). The coup de la glotte was a stroke of genius in that it provided the practical means for achieving a bright and efficient tone quality.

The importance of firm prephonatory set, firm onset, and firm adduction cannot be overemphasized. When Garcia stated that the coup de la glotte was the first lesson and the basis of his teaching, he was acknowledging that pupils did not normally use firm glottal closure until they were taught to do so. The "normal" glottal setting for conversational speech and many types of singing is loose glottal closure. Firm phonation is an "unnatural" adjustment of the vocal folds. I suspect it is rarely used outside of operatic and concert singing, and even then it is not used consistently. Its use requires a special type of training -- one that is alien to the young voice student. In my nearly forty years of teaching undergraduate voice students, I do not remember even one student who sang with firm onset and firm phonation without being trained to do so. It is just too "unnatural." From the time we are children, we are told that singing is an extension of speech. To a degree, this may be true of some types of singing. But classical singing techniques engage the vocal instrument in altogether more vigorous, complex, and unnatural ways. As someone once said, "Good singing is to speech what dancing Swan Lake is to walking around the block." I reiterate that firm onset and firm phonation must be taught. What is disappointing to me is the number of singing students who study for years, but never get that first lesson, and never reach their potential as singers. Of course, the widespread use of the microphone has also worked against the Garcia vocal ideal, since it requires little or no acoustic energy among the high-frequency components of the tone -- that is, the voice does not have to "project" like the trained voice.

To many voice teachers in Garcia’s time, the coup de la glotte was not a stroke of genius, but of madness. Despite Garcia’s frequent attempts to describe it as firm glottal closure leading to firm phonation, his critics insisted on interpreting it as a glottal plosive, that is, a prephonatory set with a tight glottis and elevated subglottal pressure, and with an onset that suddenly lets the air burst through the glottis with a certain violence that couldn’t be very good for the vocal folds. This misinterpretation of Garcia plagued him all his life.

But Garcia also had his supporters, including Matthilda Marchesi, a famous teacher in her own right, and Matthilda’s daughter Blanche Marchesi, who defended Garcia’s teaching with passion. Other disciples included Julius Stockhausen, Max Friedländer, Hermann Klein, and many others. As well, many of Garcia’s pupils became famous singers, including his sisters Maria Malibran and Pauline Viardot, Johanna Wagner (Richard’s niece), Jenny Lind, who was known as the "Swedish Nightingale," and too many more to list here (Stark 1999, 6).

The Garcia controversy went on to include many doctors and physiologists who made their own observations with the laryngoscope and drew their own conclusions, despite the fact that they were non-singers themselves and had little insight into the singing voice (Stark 1999, 78-80, passim). Morrell Mackenzie, in his widely read The Hygiene of the Vocal Organs of 1886 -- a book that went through at least seven editions -- called the shortened glottis "stop-closure," but said it only occurred in high notes (Mackenzie 1890, 56-57, 257-277). In 1941 and 1942 Joel Pressman found that in singing high notes, vocal fold vibration was restricted to the anterior two-thirds of the glottis. This is sometimes called Pressman’s "damping factor" (Pressman 1941, 1942). (What an apt name "Pressman" is!) Other researchers who discussed the shortened glottis include Willard Zemlin (1968, 195), Richard Luchsinger and Godfrey E. Arnold (1965, 80), and Wilbur J. Gould (1977). Most of these authors were medical doctors and/or laryngologists who viewed the glottis with a laryngoscope, and associated the shortened larynx with high notes, but Gould speculated that the shortened glottis was associated with high subglottal pressures and a low larynx. None of these authors went as far as Garcia, who maintained that the shortened glottis could be present not merely on the high notes but throughout the vocal range. Gould and Hiroshi Okamura, in a 1974 article, asserted that the prephonatory setting of the arytenoids is "the crucial period during which the entire character of phonation may be determined" (Gould and Okamura 1974, 359). This was precisely Garcia’s point with his coup de la glotte.

Despite his success as a teacher and author, Garcia’s method ignited a heated controversy that has raged to this day. I have described and documented this controversy at length elsewhere (Stark 1999, 14-20), so I will merely summarize it here by saying that Garcia and his disciples were viciously attacked for ruining voices with the use of the coup de la glotte. One of his chief adversaries was a Dr. Henry Holbrook Curtis of New York who wrote: "The shock , or coup de glotte [sic], is death to the voice; it is born of ignorance, and to teach or allow its continuance is a crime. We have no words strong enough to condemn it" (Curtis 1909, 159). Curtis recommended an aspirate onset with a loose throat, and gave exercises "for Acquisition of Relaxed Throat" (Curtis 1909, 175-206). Blanche Marchesi, a Garcia stalwart, replied "War was declared," and referred to Curtis’s followers as "ignoramuses" (B. Marchesi 1932, 91-93). (These authors were not nearly so polite as we are!) By the early 20th century the two camps became known as the "local effort" school and the "no-effort" school respectively, and even George Bernard Shaw, whose mother was a voice teacher, entered the fray against Garcia and the "coup de glottists" (Shaw 1932, 2:209-214, 248-9).

A new wrinkle regarding glottal closure occurred in the 1950s and 1960s, when some researchers became infatuated with the aerodynamic principle known as the "Bernoulli effect," a principle in fluid dynamics in which the pressure in a stream of fluid is reduced as the speed of the flow is increased. In 1958, Janwillem van den Berg of Groningen stated that the Bernoulli effect was also at work during phonation (van den Berg, 1958). His theory stated that the air stream in the trachea has a constant velocity until it reaches the narrowed glottal restriction. As the air traverses the glottis, the velocity increases, thereby creating a negative pressure between the vocal folds that sucks the folds together in the closing phase of the vibratory cycle. Van den Berg proposed a "myoelastic-aerodynamic" theory of voice production in which vocal fold vibration was explained as a combination of muscular and Bernoulli forces (van den Berg 1958).

William Vennard, a well-known California voice teacher, author, and researcher (he worked with van den Berg in Groningen, as well as with other researchers) was more than a little fascinated with van den Berg’s explanation of the Bernoulli effect in singing. He wished to put it to good pedagogical use, and suggested that "flooding the tone with breath," and beginning the tone with an aspirate, an imaginary [h], or even a sigh, would draw the vocal folds together by means of the Bernoulli effect (Vennard, 1961; 1967, 211-212). In an article titled "Coup de Glotte: A Misunderstood Expression," Vennard and Nobuhiko Isshiki tried to explain Garcia’s coup de la glotte with reference to the Bernoulli effect. They reported that when they used the "imaginary h" as a means of beginning the phonation with the breath, they came

nearer meeting the condition described by Garcia for the coup de glotte than any other attack in this study. There is a flow of breath for a quarter of a second before the tone begins, but the ear does not detect the aspirate sound because the rate is lower and the flow is smooth. This momentary flow sucks together the vocal folds so smartly that there is a sensation of the breath striking them ... it is the belief of the authors that this is the attack that Garcia had in mind when he spoke of the coup de glotte" (Vennard and Isshiki 1964, 17-18).

As we have seen, this is not what Garcia had in mind at all. Neurologist Barry Wyke maintained that the motor system that controls prephonatory set can be programmed to reduce the time interval between prephonatory tuning and phonation to as little as 50 milliseconds in trained singers (Wyke 1980, 46-47). This is not even close to the one quarter of a second of aspiration advocated by Vennard and Isshiki, but it is consistent with Garcia’s coup de la glotte. Vennard and Isshiki, in their zeal to legitimize the Bernoulli principle in singing, "shoehorned" Garcia’s theory to fit their own misconceptions. The influence of Vennard’s breathy onset was widespread, and his misinterpretation of Garcia’s coup de la glotte was another setback for Garcia’s theories.

More recently a 1982 study by Ronald Scherer and Ingo Titze stated, "The ‘Bernoulli effect’ of negative pressures within the larynx, which has been thought to play a major role in laryngeal vibration, must be reevaluated empirically. There is little doubt that the effect exists, but our results indicate that in magnitude it may not be as large as formerly predicted" (Scherer and Titze 1982, 79). It can be argued that, since the wider glottal opening reduces the velocity differential between the trachea and the glottis, the Bernoulli force is actually reduced. Other factors, such as the elastic recoil forces of the vocal folds, play a far greater role in vocal fold adduction. In his 1994 book, Ingo Titze provides an analysis of self-sustained vibrations, in which he notes that the Bernoulli forces cannot distinguish between the inward and outward movements of the vocal folds, and that "the mechanism for continued energy transfer from the airsteam to the tissue forces involves more than the Bernoulli forces alone" (Titze 1994, 80-82). All these objections indicate that the Bernoulli effect’s influence on glottal closure seems to have been exaggerated, and we are back to the primary importance of the myoelastic function in singing. However, there is still a lot of sighing going on in voice studios everywhere. This is a sad reflection on an idea gone wrong.

The matter of glottal closure and its effects on voice quality remains a central issue in voice science today, as well as in vocal pedagogy. In its new manifestation, glottal closure has become associated with the concepts of "breathy phonation," "flow phonation," and "pressed phonation," depending upon the strength of adductive forces as well as vertical laryngeal positions. Let me throw out a few ideas that may stimulate further thought and research in this area.

To begin, I’m sure we can all agree that breathy phonation is not desirable in a classically trained singer. It is inefficient; it does not create high-frequency components in the voice; and it requires a microphone, which is unsuitable on the opera stage or recital hall. (Recently Francis Robinson of the Metropolitan Opera declared that the Met has never used, and never will use microphones.) But when it comes to flow vs. pressed phonation, there is confusion and disagreement. One of the problems is that laboratory measurements take short samples, concentrating on just one type of phonation. But when one listens to great singers, one hears a variety of voice qualities that include both flow and pressed phonation, with every degree in between, depending upon the expressive requirements of the music. This is what Garcia referred to as "all the tints of the voice." A singer does not sing a "revenge" aria with the same tone quality as a lullaby. The use of mezza voce (half-voice) precludes the use of pressed voice, as do the high, soft tones that some singers are able to execute. To cite one singer whose voice will be familiar to most of you, a famous tenor from the 1950s by the name of Mario Lanza made a career of moving from loose phonation to firm phonation in the same song, as a means of increasing the emotional impact of the piece as it progressed toward a stentorian conclusion, especially in songs like "The Loveliest Night of the Year." But listen to his "Vesti la giubba" from Leoncavallo’s I Pagliacci, and I swear you will hear no loose phonation. This brings up another point. Have you ever noticed how "crossover" singers usually seem to cross from operatic singing to "pop" singing, and rarely the other way around? The classically-trained singer can easily croon, but the crooner cannot sing opera.


Regarding the definitions of flow and pressed phonation, Johan Sundberg wrote, "Pressed phonation is characterized by a high subglottic pressure combined with a strong adductive force" (Sundberg 1987, 80). Rolf Leanderson and Sundberg use the term "adductive force" to refer to the ratio between subglottal pressure and transglottal air flow during phonation (Leanderson and Sundberg 1988, 9). Sundberg says that pressed phonation seems to be "affiliated with a raised larynx position," and both he and Leanderson maintain it is unhealthy (Sundberg 1987, 92; Leanderson and Sundberg 1988, 8-9). Ingo Titze also favours flow phonation as opposed to pressed phonation. He defines pressed voice as "tight adduction," breathy voice as "loose adduction," and optimal voice as near the centre, but barely on the breathy side (Titze 1994, 227). He says, "A pressed voice is usually more efficient (aerodynamically and acoustically) than a slightly breathy voice ... but what price is paid for forceful adduction? (Titze 1994, 245). He maintains that "pressed voice carries with it a higher risk of damage," and, like Sundberg, recommends the flow mode of phonation (Titze 1994, 249). Like Vennard, he advocates the use of "silent [h] as a means of weakening the adductory force (Titze 1994, 91-108).

I believe that what must be included in the description of pressed phonation or flow phonation are the specifics regarding subglottal pressures, the position of the arytenoids, the relative length of the vibrating portion of the glottis, airflow rates, closed quotients, and laryngeal height, all of which are necessary parts of the equation. I agree that singing with strong adductory forces and a high laryngeal position is not desirable and may be dangerous (although there are plenty of "belters" who have been getting by with it for years!). My problem with recent characterizations of pressed phonation is that this is not how trained singers sing! There is general agreement that well-trained singers sing with a lowered larynx, but in a skilled way that is sufficient to create a desired voice quality.

Thomas Shipp offered an argument against the notion that pressed phonation is equivalent to vocal hyperfunction. Unlike Sundberg and Titze, Shipp took into account the effect of a lowered larynx on pressed phonation. He wrote, "If laryngeal elevation does, indeed, facilitate vocal fold adduction such as during swallowing, a theory of opposites would argue that lowering the larynx would decrease this adductory force. Such a gesture would inhibit the singer’s hyperadducting the folds and producing ‘pressed’ or ‘tight’ phonation" (Shipp 1987, 219). Trained singers who sing with a lowered larynx thus compensate for strong glottal closure and high subglottal pressure.

In finishing, I ask your indulgence in letting me describe to you some of the measurements of my own voice, taken repeatedly over many years at the Groningen Voice Research Lab. I sing with a shortened glottis, the arytenoids being always pressed together. The shortened glottis also becomes a narrowed glottis, with the vocal folds remaining only narrowly separated during the open phase of the vibratory cycle. My subglottal pressures range from 40 to 60 cmH2O when singing forte in the upper range, which I believe can be considered moderate for a lyric tenor. My airflow rates range from 70 to 100 ml/sec., which can be regarded as very low. My closed quotients with the customary pinched glottis range from 60 to 75 percent, and on occasion over 80 percent, which can be considered large to very large. None of this could be associated with a relaxed throat or with flow phonation. In most ways these figures lean toward the current descriptions of pressed phonation and vocal hyperfunction. However, I also modulate the vertical laryngeal position in order to control the balance of muscular forces. I prefer to say that I sing with chiaroscuro, and with vocal efficiency, not with pressed voice. Incidentally, my larynx is healthy, and I have never had vocal polyps, nodules, contact ulcers, or any other vocal pathology. While all of this is anecdotal, it may at least cause us to look more closely at the matter of glottal closure and voice quality.

I want to end by returning to the quote I started with: "The art of singing has achieved a high level of refinement and sophistication -- not so the science" (Rubin, LeCover, and Vennard 1960, 29). Part of my task as a keynote speaker was to suggest ways in which voice science might advance. In my view, there are two principal things that we should strive for. First, we should encourage world-class singers to act as subjects in our experiments. To compare amateur singers to so-called trained singers in experimental protocols is not particularly helpful, since the word "trained" can mean so many things. Second, I think that the resistance of some researchers to using commercial recordings of great singers because of the lack of laboratory controls is not wholly valid. Of course we all know that there are singers who have friends in the control booth, but there are also live broadcasts and good honest recordings that can reveal much about singing. As mentioned earlier, one of the limiting factors in voice science is that the vocal samples in the laboratory are so short, and so divorced from what goes on in real singing. But the reality of fine singing is that it is full of colours, nuances, intensity levels, the use of legato, portamento, and low pitch onset, the use of schwa, and all the other artifices of classical singing. Voice science should move toward analyzing the voice in the context of the music that makes such extraordinary demands on the voice. What we need is real singers singing real music, and we must develop better laryngoscopy, and new, less invasive instrumentation that will enable us to objectively analyze singing without scaring off our subjects.

The singing voice is a thing of wonder, and it has only given up its secrets reluctantly over the centuries. Between us all, singers, teachers, voice scientists, laryngologists, musicologists, and all lovers of good singing, we have come a long way, but of course we still have a long way to go. Let us work together to enjoy the adventure!


Bartholomew, Wilmer T. 1934. "A Physical Definition of ‘Good Voice Quality’ in the Male Voice." Reprinted in Large 1980, 299-306. [First published in Journal of the Acoustical Society of America 6: 25-33.]
Bartholomew, Wilmur T. 1945. Acoustics of Music. New York: Prentice-Hall.
Brown, William Earl. 1957. Vocal Wisdom: Maxims of Giovanni Battista Lamperti. Enlarged Edition. New York: Arno Press. [First published in 1931.]
Caccini, Giulio. 1602. Le nuove musiche. Firenze: Marescotti. [Facsimile edition New York: Performers’ Facsimiles 35, n.d.] English Translation by H. Wiley Hitchcock. Madison, WI: A-R Editions, 1970.
Curtis, Henry Holbrook. 1909. Voice Building and Tone Placing. Third Edition. London: J. Curwen & Sons. [First edition, 1896.]
Garcia, Manuel. 1847. Traité complet de l’art du chant. En deux partie, 2e édition; second partie, 1e édition. Genève: Minkoff Éditeur. [Reprint with an introduction by L. J. Rondeleux. Includes Mémoires sur la voix humaine (1840).]
Garcia, Manuel. 1855. "Observations on the Human Voice". Proceedings of the Royal Society 3: 399-408. [Reprinted in Large 1980, 123-133.]
Garcia, Manuel. 1894. Hints on Singing. Translated from the French by Beata Garcia. New and revised edition [by Hermann Klein]. London: E. Ascherberg. [Reprint edition: New York: Joseph Patelson Music House, 1982.]
Garcia, Manuel. 1975. A Complete Treatise on the Art of Singing: Part Two. The editions of 1841 and 1872 collated, edited, and translated by Donald V. Paschke. New York: Da Capo Press. [English translation of Garcia 1847, Part Two.]
Garcia, Manuel. 1984. A Complete Treatise on the Art of Singing: Part One. The editions of 1841 and 1872 collated, edited, and translated by Donald V. Paschke. New York: Da Capo Press. [English translation of Garcia 1847, Part One.]
Gould, Wilbur J., and Hiroshi Okamura. 1974. "Interrelationships Between Voice and Mucosal Reflexes." In Wyke, Barry D., ed. Ventilatory and Phonatory Control Systems. London: Oxford University Press, 347-359.
Helmholtz, Hermann. 1954. On the Sensations of Tone as a Physiological Basis for the Theory of Music. [Reprint of] Second English edition, translated, thoroughly revised and corrected ... by Alexander J. Ellis, with a new introduction by Henry Morgenau. New York: Dover. [First published in German as Die Lehre von den Tonempfindung, 1862.]
Large, John, ed. 1980. Contributions of Voice Research to Singing. Houston: College-Hill Press.
Leanderson, Rolf, and Johan Sundberg. 1988. "Breathing for Singing." Journal of Voice 2 (1): 2-12.
Mackenzie, Morrell. 1899. The Hygiene of the Vocal Organs. Seventh edition. New York: Edgar S. Werner. [First published 1886.]
Mancini. 1774. Pensieri e riflessioni pratiche sopra il canto figurato. Vienna, 1774; second edition 1777. Translated and edited by Edward Foremen as Practical Reflections on Figured Singing: The Editions of 1774 and 1777 Compared. Champagne, IL: Pro Musica Press, 1967.
Marchesi, Blanch. 1932. The Singer’s Catechism and Creed. London: J. M. Dent & Sons.
Miller, D.G., and Schutte, H.K. 1990. "Formant Tuning in a Professional Baritone." Journal of Voice 4:3, 231-237.
Miller, D.G., and Schutte, H.K. 1994. "Toward a Definition of Male ‘Head’ Register, Passaggio, and ‘Cover’ in Western Operatic Singing." Folia Phoniatrica et Logopaedica 46, 157-170. Reprinted as Chapter 7 in Miller, D.G., Registers in Singing (2000), Ph.D dissertation, University of Groningen, 125-146. 
Pressman, Joel. 1941. "Sphincters of the Larynx." Archives of Otolaryngology 33: 221-236.
Pressman, Joel. 1942. "Physiology of the Vocal Folds in Phonation and Respiration." Archives of Otolaryngology 35: 355-398.
Rubin, Henry J., M. LeCover, and William Vennard. 1967. "Vocal Intensity, Subglottic Pressure, and Air Flow Relationships in Singers." Folia Phoniatrica 19: 393-413. [Reprinted in Large 1980, 88-107.]
Scherer, Ronald C., and Ingo R. Titze. 1982. "A New Look at van den Berg’s Glottal Aerodynamics." Transcripts of the Tenth Symposium: Care of the Professional Voice 1: 74-81. New York: The Voice Foundation.
Schutte, Harm K., and Donald G. Miller. 1996. "Resonance Strategies for Tenor High Notes." Paper presented at the 25th Annual Symposium: Care of the Professional Voice. Philadelphia, 3-9 June, 1996.
Shaw, George Bernard 1932. Music in London, 1890-1894. 3 vols. London: Constable and Company.
Shipp, Thomas. 1987. "Vertical Laryngeal Position: Research Findings and Applications for Singers." Journal of Voice 1 (3): 217-219.
Stark, James. 1999. Bel Canto: A History of Vocal Pedagogy. Toronto: University of Toronto Press.
Sundberg, Johan. 1977. "The Acoustics of the Singing Voice." Scientific American (March), 82-91.
Sundberg, Johan. 1987. The Science of the Singing Voice. DeKalb, IL: Northern Illinois Press.
Titze, Ingo. 1994. Principles of Voice Production. Englewood CLiffs, NJ: Prentice-Hall.
Transcripts of the Tenth Symposium: Care of the Professional Voice. 1983. New York: The Voice Foundation.
Van den Berg, Janwillem. 1958. "Myoelastic-Aerodynamic Theory of Voice Production." Journal of Speech Research 3: 227-244.
Vennard, William. 1967. Singing: the Mechanism and the Technic. Revised edition, greatly enlarged. New York: Carl Fischer. [First edition, 1949.]
Vennard, William. 1961. "The Bernoulli Effect in Singing." Bulletin of the National Association of Teachers of Singing (February): 8-11.
Vennard, William, and Nobuhiko Isshiki. 1964. "Coup de Glotte -- A Misunderstood Expression." Bulletin of the National Association of Teachers of Singing (February): 15-18.
Wyke, Barry D. 1980. "Neurological Aspects of Phonatory Control Systems in the Larynx: A Review of Current Concepts." Transcripts of the Eighth Symposium: Care of the Professional Voice 2: 42-53. New York: The Voice Foundation.
Winckel, Fritz. 1952. "Physikalische Kriterien für die objektive Stimmbeurteilung." Folia Phoniatrica 5: 232-252.
Zacconi, Lodovico. 1592. Prattica di musica utile et necessario si al compositore. Venetia: Girolomo Polo.


Professor of Music, Mount Allison University
Sackville, New Brunswick, Canada


Last update, Nov 15, 2002      Webmaster: Harm K. Schutte             
Stay informed by checking regularly this web site
(November 2002, photographs of the conference added)