mammals, the source is typically the larynx. In birds, an evolutionarily novel
structure called the syrinx serves as the voice source. In both cases, the
source contains mobile elastic structures that act as mechanical vibrators
and can reduce or stop the passage of air through the source by constricting its lumen. In the larynx, these vibrators are the vocal folds, sometimes
called vocal cords. In the syrinx, the identity of the vibrators was long
thought to be the medial tympaniform membranes (Miskimen 1951;
Greenewalt 1968; Gaunt and Gaunt 1985). However, more recent direct
visualization via endoscopy suggests that the vibratory structures are the
syringeal labia in passerines (Goller and Larsen 1997b) and the lateral tympaniform membranes in pigeons (Goller and Larsen 1997a). Although the
medial tympaniform membranes may play some acoustic role, their complete ablation does not prevent nearly normal vocalization (Larsen and
Goller 1999). In both the larynx and syrinx, energy created by the passage
of air through the constriction between the vibrators acts to set them into
motion. When the vibrators collide (or approach close enough to modulate
air flow), they generate acoustic energy. The main acoustic difference
between the larynx and syrinx is their location: the larynx is located at the
top of the trachea, whereas the syrinx is located at its base. Although birds
also possess a larynx, there is little evidence that the avian larynx is used as
a sound-producing source (see White 1968 for a possible exception).
There is a long history of scientific attempts to understand the functioning of the human voice source, starting with Johannes Müller’s pioneering
work with excised larynges (Müller 1848).The further work of van den Berg
(1958, 1968) and Titze and colleagues (Titze 1994) has deepened this understanding, and we now have a detailed and accurate model of vocal production at the voice source, which is called the “myoelastic–aerodynamic
theory.” The vocal folds act as mechanical vibrators, with their own elasticity and mass, that are coupled to the aerodynamic flow from the lungs to
generate self-oscillation. Before oscillation can begin, the vocal folds must
be placed in an appropriate “phonatory” position, closer to one another
than during normal breathing. Once this position is attained, air flow from
the lungs can set up sustained oscillations. The critical factor is that energy
must be pumped into the system in phase with the vocal fold oscillation:
there must be a greater pressure pushing the folds apart as they are opening
than while they are closing. In human speech, this is made possible by a
change in the vertical geometry of the folds, with a divergent glottis during
closing leading to less force than the convergent glottis during opening.
Although it is frequently stated that the Bernoulli force is adequate to
maintain oscillation, this alone is not adequate because it is equally strong
during opening and closing (Titze 1976, 1980). Similar considerations must
apply in all tetrapod sound sources; Titze (1994) is recommended as an
excellent introduction to the topic.
The quantitative details of laryngeal vibration are still an area of active
investigation, even for humans. Investigations of nonhuman larynges
3. Unpacking “Honesty”
75
structure called the syrinx serves as the voice source. In both cases, the
source contains mobile elastic structures that act as mechanical vibrators
and can reduce or stop the passage of air through the source by constricting its lumen. In the larynx, these vibrators are the vocal folds, sometimes
called vocal cords. In the syrinx, the identity of the vibrators was long
thought to be the medial tympaniform membranes (Miskimen 1951;
Greenewalt 1968; Gaunt and Gaunt 1985). However, more recent direct
visualization via endoscopy suggests that the vibratory structures are the
syringeal labia in passerines (Goller and Larsen 1997b) and the lateral tympaniform membranes in pigeons (Goller and Larsen 1997a). Although the
medial tympaniform membranes may play some acoustic role, their complete ablation does not prevent nearly normal vocalization (Larsen and
Goller 1999). In both the larynx and syrinx, energy created by the passage
of air through the constriction between the vibrators acts to set them into
motion. When the vibrators collide (or approach close enough to modulate
air flow), they generate acoustic energy. The main acoustic difference
between the larynx and syrinx is their location: the larynx is located at the
top of the trachea, whereas the syrinx is located at its base. Although birds
also possess a larynx, there is little evidence that the avian larynx is used as
a sound-producing source (see White 1968 for a possible exception).
There is a long history of scientific attempts to understand the functioning of the human voice source, starting with Johannes Müller’s pioneering
work with excised larynges (Müller 1848).The further work of van den Berg
(1958, 1968) and Titze and colleagues (Titze 1994) has deepened this understanding, and we now have a detailed and accurate model of vocal production at the voice source, which is called the “myoelastic–aerodynamic
theory.” The vocal folds act as mechanical vibrators, with their own elasticity and mass, that are coupled to the aerodynamic flow from the lungs to
generate self-oscillation. Before oscillation can begin, the vocal folds must
be placed in an appropriate “phonatory” position, closer to one another
than during normal breathing. Once this position is attained, air flow from
the lungs can set up sustained oscillations. The critical factor is that energy
must be pumped into the system in phase with the vocal fold oscillation:
there must be a greater pressure pushing the folds apart as they are opening
than while they are closing. In human speech, this is made possible by a
change in the vertical geometry of the folds, with a divergent glottis during
closing leading to less force than the convergent glottis during opening.
Although it is frequently stated that the Bernoulli force is adequate to
maintain oscillation, this alone is not adequate because it is equally strong
during opening and closing (Titze 1976, 1980). Similar considerations must
apply in all tetrapod sound sources; Titze (1994) is recommended as an
excellent introduction to the topic.
The quantitative details of laryngeal vibration are still an area of active
investigation, even for humans. Investigations of nonhuman larynges
3. Unpacking “Honesty”
75