Paul Cook Voice

This research paper examines the extent to which scientific research on breathing since 1970 has informed the art of vocal teaching.

breath mechanism
‘In the clinic, I see many singers unaware of how much excess energy they are expending on breathing. When asked about their understanding of breathing for singing, they parrot phrases they have heard (e.g., “we should breathe with our diaphragm”), but admit to confusion about the actual mechanics of breathing.’ – Michael, D. (2010), Journal of Singing, Vol. 66, No. 5, p. 547

Critically examine the extent to which scientific research on breathing undertaken since 1970 has informed the art of vocal teaching.

by Paul Cook

Of the circa fifty articles related to scientific research in the Journal of Singing, over three quarters of them have been written within the last twenty years. This suggests that scientific research into singing has been a relatively recent phenomenon and, when one considers the ever-evolving advances in medical technology, this might be a sensible conclusion.

However, serious scientific research began several decades before the end of the last century. Steinhauer et al. (2017, pp.10-46) documents how Jo Estill spent much of her adult life collaborating with elite researchers in the field of medicine and speech science in an attempt to bring knowledge to the mystery surrounding human speech and singing. From 1972, Estill spent several years working as a research associate as part of a team of eminent researchers in a variety of locations including five years at the State University of New York. She went on to co-write many articles and papers which were published around the world in academic journals and presented at symposia.

However, even science that has been widely revered for several decades can always be subject to peer review and that nothing is infallible. Despite the findings of scientific research undertaken by Estill still being used today by hundreds of teachers around the world, Kayes (2017) casts doubt on the reliability of some of the most important elements of the model, most notably cricoid tilt in belt quality. Kayes argues that it is physically impossible to tilt the cricoid during belt and gives a comprehensive explanation about how Estill misinterpreted some of what she believed she discovered and that some of her research was based on assumptions which now, some thirty years on, are irrefutably unfounded.

Simonson (2014, 2015) provides a small insight into some of the diverse topics that have been the focus of researchers in recent years which include studies into vocal tract shapes, the physiological acoustic characteristics of the female music theatre singer and differences in the postures of dysphonic and normal speakers. It seems that no area of vocal science is now inaccessible to today’s researchers with access to modern technology with significant advances made in neuroscience to examine brain function and singing (Peretz et al. 2004; Maxfield, 2019), instruments that can display the behaviour of the larynx thus removing the need to insert a fibre optic camera into the patient and detailed spectral-acoustic analysis of different voice types.

Modern research into breath methodology is constantly evolving into a diverse area of enquiry with a higher level of expertise due to ongoing scientific advancements and development in technology (Cleveland 1992). A relevant example of recent research into breathing using technology, in this case, lung function (Cowgill 2009) used anthropometric somatotyping to select thirty participants with the objective of comparing lung capacity and breathing patterns of different body types (ectomorphs, mesomorphs and endomorphs). Each participant was videoed singing the first four lines of ‘Caro Mio Ben’. The expansion of the lateral chest wall and vertical movement of the thorax of each participant was subsequently measured on playback before being measured by a spirometer to measure the rate of air flow and volume of the air expelled from the lungs.

A Very Brief Overview of the Mechanics of Breathing

While a large number of concepts about breath control exist, (Emmons 1988), explanations of the breathing cycle are readily available from a variety of sources which broadly agree with each other (Dimon, 2018; Miller, 1996; Chapman, 2006; Taylor, 2008).

At rest, the breath cycle lasts approximately four seconds, the in breath lasting around one second and the out breath lasting the remaining three (Miller 1996, 20). The cycle is longer during performance breathing due to the breath being managed by the singer to perform sustained, longer phrases.

Chapman (2006, 48) notes that the primary and secondary muscles of inspiration control the increase in space in the thoracic cage as the lungs fill and expand with air. Chapman identifies the intercostal muscles and the diaphragm (which contracts downwards) as being the primary muscle group of inspiration and the secondary muscle group as muscles that surround the rib cage including those in the neck, shoulders and arms.

Taylor (2008, 20) succinctly notes that exhalation is caused by the abdominal muscles pushing back so that the diaphragm releases upwards consequently forcing air out of the lungs and to the vocal folds causing them to vibrate and produce sound.

Where to Place the Breath?

Estill (2005) notes that breathing is dynamic, that the respiratory system is governed by the involuntary nervous system and is, consequentially, subject to constant change. Estill further observes that whilst, tidal breathing is involuntary, performance breathing is voluntary and that neither preference or ‘attractor state’ for belly or chest breathing is superior to the other.

Chapman (2006,) contradicts this view and argues that chest breathing is inefficient because the thoracic space gained is small, the muscles engaged are primarily for postural alignment and the subsequent, increased tension affects vocal tone, resonance and quality negatively.

McCoy (2010) suggests that breathing is widely seen as a solution for a wide variety of vocal problems however, in his view, poor breath management is often the cause of common bad habits including incorrect laryngeal positioning and jaw tension.

Emmons (1988) comments that there are “hundreds of concepts” concerning breath control in singing and that psychological perceptions are often not representative of physiological reality. For example, ‘sing with the back,’ cited in Leanderson and Sundberg (1988) as a phrase often used by singing teachers, is a perception of where the physical effort of breathing is occurring because inhalation causes an expansion of the lower rib cage thus causing a sensation in the dorsal region. Miller (1996, 278) condemns pedagogy that falsely advocates the movement of specific muscles, in the back in this example, which are not essential to the physiology of breathing.

There is no shortage of breathing exercises: (Chapman, 2006, Kayes, 2004, Williams, 2019) providing a wide range of breathing exercises which differ widely on their focus and purpose. A seemingly inexhaustible plethora of videos exist online, each one often presenting their own methodology and imagery. Practising breathing exercises in isolation does not necessarily guarantee an improved vocal technique but masterful breathing, not masterful singing (Steinhauer, Klimek and Estill, 2017, p. 38).

Kayes (2007) suggests that singers are often so obsessed with breath control that they do not breathe out enough and/or that they are regularly too tense to inhale efficiently. Excluding fixation on breath control and tension, Kayes argues that the only other breathing issues experienced by singers relates to a misconception about how to breathe and manage the breath.

In contrast to the obvious importance placed on breathing by researchers, eminent teachers of singing and authors, the lack of emphasis on breathing in the extensive research undertaken by Jo Estill is perhaps provocative to some practitioners. Breathing, referred to as ‘power’ in the Estill Voice Model is seldom mentioned and no specific exercises or figures exist within in its extensive training structure for aspiring teachers. Speed (2012) argues that the model is fundamentally concerned with the behaviour of the source (vocal folds) and the conditions of the filter (vocal tract) and that the breath should be appropriate to the task to produce clear tone. Even when breath is purposefully mixed into the sound, for example in falsetto quality, the model does not address how the breath is produced but only how it is managed at the source. Indirectly, Indik (2009) claims that breathing exercises, even those that remove phonation, are effective and retain relevance. However, he then goes on to suggest that exercises that ignore the behaviour of the breath upon meeting the vocal folds can lead to misconception. Douglas and Yarnell (2005) are more aligned to Estill’s views and suggest that effective breath management requires simultaneous respiratory muscular balance to create a flow of breath of sufficient volume to meet the requirement of the singer.

Freed (1994) identifies widespread disagreement amongst pedagogues about teaching singers how to breathe and highlights the significant variance in how terminology is understood and applied.

Mouth or Nose?

Miller (1997) notes that breathing through the nose in activities unrelated to singing is better for humans because the air is filtered and warmed before reaching the lungs. Inhaling through the mouth does not provide an opportunity for the air to be filtered and there is less distance for the air to travel before reaching the lungs, hence it is colder. He also observes that breathing in through the nose takes longer than through the mouth which is an obvious consideration for the performer who may have only a very short rest between phrases. Rapid in breaths therefore require inhalation through the mouth. However, Miller argues that a breath inhaled through the nose is usually silent and a more complete breath. He concludes that breathing in through the nose during extended gaps between phrases is more desirable than breathing through the mouth. He also suggests that every singer should determine to learn how to inhale silently through the mouth.

Estill (2005) notes that if the false vocal folds are constricted during inhalation then the breath is disturbed and therefore not silent. Constriction of the false vocal folds during phonation can cause disruption to the vibration of the true vocal folds and rate of air flow. Estill acknowledges that the muscles that control the false vocal folds were, at the time of writing, unknown but suggests that the posture of complete retraction of the vocal folds can best be imagining the sensation of crying or laughing.

The Concept of Support

Williams (2019, 91) suggests that the term ‘support’ is widely misunderstood and misappropriated. Many singers, she argues, often associate support with simply gripping their abdominal muscles which does not help efficient phonation. In her opinion, support is more to do with maintaining balance and stability of the out breath and, consequently, the sound.

Macdonald et al. (2012) describes the balancing effect more fully and observes that the recruitment patterns and thickness of the transversus abdominus muscle and the internal oblique muscle play a role in supporting the healthy voice. By using ultrasound technology, the study found that different sections of the diaphragm do not necessarily move simultaneously with the supporting muscles and that the movement is greater in the posterior because this is primarily where the majority of muscle fibres are located. However, the anterior muscles are used by healthy singers primarily to position the abdominal wall and to manage the subglottal pressure.

Emmons (1988) notes that the diaphragm is passive and that the thoracic muscles control the rate of breath from the lungs. Kayes (2007, 27) agrees that one cannot support from the diaphragm alone since it has no nerves endings and cannot be moved independently Emmons states furthermore, continuous adjustment in volume/capacity and pressure is required to accommodate performance requirements such as voice quality. No muscle moves alone and Emmons notes that the proficient singer must be able to balance the unceasing fluctuation between the abdominal and thoracic muscles during exhalation to ensure managed and effective equilibrium of pressure and air flow.

Kayes suggests that support can mean both regulating the breath from below the larynx and also the recruitment of other muscles to assist work in the vocal tract. The latter, referred to in the Estill Voice Model as ‘anchoring’, is based upon the assumption that the vocal folds themselves sometimes are not sufficiently strong to manage the pressure or forces of the out breath (for example in belt quality) and that muscular recruitment has to come from elsewhere to provide support to avoid creating tension in the larynx. Kayes refers to the relationship between the larynx and other recruited muscles as the ‘voice-body connection.’ Kaye’s (2007, 75) definition of what she calls ‘vocal tract anchoring’ is uncannily similar to that of Jo Estill. The larynx, being suspended, has little or no stability or muscular support against which to push during high energy breath. By altering the position of the head and neck muscles, the soft tissue and muscles surrounding the oral cavity and larynx become harder to provide a sense of extra space (for resonance) and support. In torso anchoring, the intercostal muscles are recruited to help bring stability to the back of the torso (and consequently the breath) during more active singing.

The lungs

Although the lungs are the most important respiratory organs, they do not move of their own accord but are subject to the movement of the thoracic cage musculature to which they are attached (Miller, 1996). We do not alter the air as such when we breath in and out (Dimon, 2018). Instead, the muscles of the thoracic cage expand the lungs thus increasing the space within them and air is sucked in as a result: the opposite occurs with the out breath.

Estill (2005) notes that lung capacity is determined by the space available in the thorax which can vary slightly according to physical stature. Cowgill (2009) conducted a study which compared the lung capacity of mesomorphs, ectomorphs and endomorphs who were engaged in formal vocal training and found that the range of the mean capacity was only 0.25 of a litre.

Williams (2019) highlights the implications for the vocal teacher when teaching young people with smaller thoracic space available. The lungs, whilst adopting the upright position of the adult by the age of 8 years old, do not reach full adult maturity until after puberty, sometimes as late as 20 years of age. The consequence of the relatively late development of lung capacity means that adolescent students are often unable to sing long, sustained phrases.

Although the lungs are passive and do not move of their own accord, the elasticity of their composition, in conjunction with the rib cage helps to create a recoil force. Leanderson and Sundberg (1988) describe how the lungs act like rubber balloons and create a passive expiratory force to expel the air inhaled. The force increases with larger amounts of air. After maximum exhalation, the lung recoil force, regulated by pressure in the pleural cavity (the fluidic space which separates the lungs from the thoracic cage) forces the diaphragm into the rib cage.

Dimon (2018) notes that the maximum amount of breath that can be taken in by the lungs of an adult is typically six or seven litres. The residual volume typically left after forced exhalation (or performance breathing in the Estill Voice Model) is approximately two litres. How to access this two litres of residual lung capacity, when the brain is convinced that the lungs are empty has long since posed questions for scientific researchers. Titze (1985) argues that a neurological ‘hook up’ between the lungs and larynx exists which implies that the brain can override the urge to inhale before it is necessary. Indik (2009) suggests eleven strategies for prolonging the breath at the point of perceived maximum exhalation, but the implications and efficacy of these would demand a level of enquiry which a study of this brevity would be unable to address. McCoy (2014) suggests that singers simply contract their expiratory muscles with maximum force to access the last molecules of breath.

Posture and Breathing

Berry (2008, 20) makes the point that tension causes more tension. He suggests that tension can cause a reduction in breath capacity. If the spine is not completely straight then the rib cage is unable to open properly. Berry stresses the need for a relaxed posture, particularly in the neck to avoid tension in the larynx: relaxation and breathing are completely dependent on each other. Taylor (2008, 39) also supports the view about the importance of spinal alignment, commenting that problems often start if the hips are too far forward or back.

Williams (2019, 88) stresses the importance of avoiding raising the chest when inhaling because the action of lifting the ribs can cause tension in the neck and shoulders. Instead, she suggests, that breathing should be low in the belly and that focusing on a sensation of widening the back and lengthening the spine promotes the correct posture for breathing.

Chapman (2006, 25) argues that when considering posture for breathing, the breath management, core stability (using the abdominal musculature) and posture should be considered simultaneously because they are all interdependent. She notes that an overly upright posture encourages high chest breathing which can cause rigidity in the upper body and undermine the vocal function. Chapman concludes that core stability is vital for effective posture when breathing and flexibility must be maintained in the rib cage and abdominal muscles to allow an efficient, unimpeded breath mechanism.

McCoy (2010) acknowledges that there are many different approaches to breathing posture but that common features exist amongst the majority of them, most notably the need to keep the upper chest and shoulders free from tension upon inhaling. He also refers to high chest breathing as a potential source of distorted phonation.

Brief Overview of Some Notable Methods of Breathing Used by Teachers and Singers

Miller (1991) suggests that many methods of controlling the breath exist and that they are seldom completely in agreement. He cites the centuries old appoggio method (which literally translates as ‘support’) as a well proven means of establishing balance between the breath and the musculature that supports it. Miller also acknowledges, but does not name, some methods that completely ignore how breath is managed below the larynx. Perhaps Miller is alluding to the Estill Voice Model which according to Speed (2012) is widely perceived as not considering the management of the breath subglottis at all until it comes into contact with the true vocal folds.

The Appoggio Method

Chapman (2006, 40) suggests that, during the last few decades, there has been widespread confusion about the application of the term “appoggio”. She argues that by misapplying the term to the manner in which the muscles of inspiration and exhalation work together, some pedagogues have not taken into account the interdependence of the diaphragm, rib cage and abdominal muscles.

Miller (1991) suggests that the appoggio method is perhaps the most efficient system of breathing for singers and also the most widely used. He describes the method as a means of achieving and maintaining a dynamic balance between the abdominal and torso muscles during singing. This action inhibits the rate and pressure of the out breath giving the singer the ability to sing longer, more sustained phrases.

Miller points out that the appoggio method can be taught to all singers but that the morphology of the student must be considered. Upon inhalation the singer is aware of expansion of the lateral and anterior muscles and also the lower dorsal region at the bottom of the rib cage. He observes that tenors often have shorter rib cages than basses and women usually have longer rib cages than men. Rib cages in females are also lower than in males which means that the expansion will occur lower in the torso.

The potential implication for the teacher, striving to teach appoggio breathing, Miller suggests, is that they may not always perceive that the muscular expansion during inhalation is occurring in the right place. For example, the chest of a bass with a longer rib cage is likely to move less, when expanding his lower dorsal muscles, than that of a tenor who has less space between the ribs. Miller concludes that it is best practice to locate the bottom of the student’s rib cage before teaching appoggio. Teachers should also refrain from using their own body as an example because they may have a different physique than that of the student.

The Accent Method

Widely used on the continent, the Accent Method was devised and developed by Svend Smith who was a Danish phonetician (British Voice Association, 2018). The method was conceived originally as a rehabilitation therapy system and is used primarily to help improve the co-ordination of breathing and phonation with an emphasis on resonance and articulation. According to the BVA, the method is one of the few breathing systems that has a good evidence base following thorough research.

The Alexander Technique

The Alexander Technique is a method that focuses on retraining the postural habits of an individual (Hudson 2002). It has been used widely by singers to release tension, performance anxiety and muscular misuse.

Hudson explains that Alexander believed that many breathing exercises used by singers to improve their breath control focused too much on expanding the chest and abdomen and that other muscles that should be recruited to maintain a sense of balance and equilibrium were contracting, particularly in the lumbar region. He believed that many breathing exercises encouraged habitual lowering of the chest during exhalation which potentially creates pressure in the thorax, a misalignment in the centre of gravity and an unnecessary depression of the larynx. Alexander believed that to successfully achieve an efficient posture for breathing that the muscles of the head and neck should be relaxed, that the head should be in a forward position and upward alignment and that the torso should be allowed to widen and lengthen.


It is clear that the scope and diversity of scientific research into breathing is profound and widespread. Furthermore, the vast resources into breathing methodology and research findings found in respected literature on singing pedagogy for example (Chapman, 2006; Kayes 2007; Miller 1996) is indicative of how endemic and embedded scientific study has become in the pursuit of informing vocal teaching. In addition to the literature mentioned above, professional bodies such as The National Association of Teachers of Singing and The British Voice Association add to the wealth of existing knowledge through the regular inclusion of new studies relating to breathing methodology in their published journals.

As technology advances and the depth of knowledge increases in the field of breathing research, previous findings have been challenged. Chapman (2006, 257)) expresses her doubts into various findings of Jo Estill’s research during the 1970s and 1980s whilst Kayes (2017) denounces the notion of cricothyroid tilt practised in the voice model as wrong.

There is significantly more disagreement among performers and singing teachers about the nature of breathing (Emmons 1988) where the perception of knowledge is often based upon imagery passed down from teacher to student rather than precise scientific research. Emmons comments that the perception of what is happening often is not representative of the physiological reality.

Some of the studies highlighted in this report exemplify how integral modern technology has become in scientific research into breathing methodology for example (Cleveland 1992; Cowgill 2009; Maxfield 2019). Further enquiry into this field of study could focus on why so many singing teachers still rely on traditional imagery and fail to embrace modern research findings.

In conclusion, there still appears to be much disagreement amongst researchers and pedagogues on the specific nature of breath management when applied to vocal technique. Miller (1991) suggests that some pedagogues hardly even consider the nature of breathing at all. (Chapman, 2006 and Williams 2019) note that even where there is a perception of understanding and agreement on a specific area of breathing methodology, that widespread misconception still exist which undermines accurate understanding.


Berry, C. (2008), Voice and the Actor, London: Virgin Books Ltd

British Voice Association (2018) The Accent Method. Available at: (Accessed: 28th January 2021).

Chapman, J. (2006), Singing and Teaching Singing: A Holistic Approach to Classical Voice, Oxford: Plural Publishing

Cleveland, T. (1992), Voice Pedagogy for the Twenty-First Century: Physiological and Acoustical Basis for Vocalises Relating to Subglottal Pressure, NATS Journal, November/December 1992, pp.25-26

Cowgill, J. (2009), Breathing for Singers: A Comparative Analysis of Body Types and Breathing Tendencies, Journal of Singing, November/December 2009, p. 144

Dimon, T. (2018), Anatomy of the Voice, California: North Atlantic Books

Douglas, J., Patenaude-Yarnell, J. (2005), Teaching Breathing, Journal of Singing, Vol. 61, No. 5, pp. 487-492

Emmons, S. (1988), Breathing for Singing, Journal of Voice, Vol. 2, No. 1, pp. 30-35, New York: Raven Press

Estill, J. (2005), Estill Voice Training Level One, Figures for Voice Control, Estill Voice training Systems International

Freed, D. (1994), Breath Management Terminology: How Far Have We Come?, Journal of Singing, Vol. 50, No. 5, pp. 15-28

Hudson, B. (2002), The Effects of the Alexander Technique on the Respiratory System of the Singer/Actor, Journal of Singing, Vol.59, No.1, pp. 9-17

Kayes, G. (2007), Singing and the Actor, London: A & C Black Publishers Limited

Kayes, G. (2017), After Estill: a conversation we need to have (video). Available at (accessed: 27 January 2021).

Leanderson, R., Sundberg, J. (1988), Breathing for Singing, Journal of Voice, Vol. 2, No. 1, pp. 2-12, New York: Raven Press

Macdonald, I et al. (2012), An Investigation of Abdominal Muscle Recruitment for Sustained Phonation in 25 Healthy Singers, Journal of Voice, Vol. 26, No.6, 2012, New York: Raven Press

Maxfield, L. (2019), Neuroscience for Singers, Part 1: Neuroanatomy, Journal of Singing, Vol. 76, No. 2, pp. 203-206

McCoy, S. (2014), On Breathing and Support, Journal of Singing, Vol. 70, No. 3, pp. 321-324

Michael, D. (2010), Dispelling Vocal Myths. Part 1: “Sing from Your Diaphragm!”, Journal of Singing, Vol. 66, No. 5, p. 547

Miller, R. (1991), In Search of the Tenth Rib, The NATS Journal, May/June 1991, p. 17

Miller, R. (1996), The Structure of Singing: System and Art in Vocal Technique, California: Wadsworth Group

Miller, R. (1997), Nose or Mouth Breathing?, Journal of Singing, Vol. 53, No. 5, pp. 37-38

Peretz, I., Gagnon, L., Herbert, S., Macoir, J. (2004) Singing in the Brain: Insights from Cognitive Neuropsychology, available at (Accessed 27th January 2021).

Simonson, D. (2014), Recent Research in Singing, Journal of Singing, Vol. 70, No. 4, p. 493-494

Simonson, D. (2015), Recent Research in Singing, Journal of Singing, Vol. 71, No. 5, p. 645-646

Speed, A-M. (2012) Breathing, Support and Estill Model, available at (Accessed: 25th January 2021).

Steinhauer, K., Klimek, M., Estill, J. (2017), The Estill Voice Model: Theory and Translation, Pennsylvania: Estill Voice International

Taylor, M. (2008), Singing for Musicals: A Practical Guide, Marlborough: The Crowood Press Ltd

Titze, I. (1985), Coordinated Breathing and Phonation, Journal of Singing, Vol. 42, No. 1, p. 22

Williams, J. (2019), Teaching Singing to Children and Young Adults, Oxford: Compton Publishing