The phonation type is seldom limited to just one phonetic segment, but on the other hand can be changed during speaking. Thus, phonation type is primarily regarded as a laryngeal setting. The laryngeal settings constitute the main set of parameters describing voice quality . Despite the articulatory movements which produce a given speech sound, the articulators can be set in some particular positions for longer stretches of speech.
In Laver's framework (1994) the supralaryngeal settings , which describe the postions of the articulators, are divided into
The supralaryngeal settings are summarized in Table 2.
The auditory effects of supralaryngeal settings manifest themselves primarily in the modification of the frequency spectrum of a produced sound. The changes affect the vocal tract formants (frequency and bandwidth of eigenresonances) and the shape of the spectrum. The supralaryngeal settings provide extra-linguistic information, for example about a speaker's habitual and social factors, but are language-dependent (the deviations from the neutral settings that are recognized as the use of a particular setting differ between languages). The effects of supralaryngeal settings are described below.
The longitudinal modifications of the vocal tract result from vertical displacements of the larynx and from the lips' protrusion or retraction . The change of vocal tract length has auditory and acoustic correlates in voice quality (section 24).
( Protruded jaw , WAV file, 18 kB)
( Retraced jaw , WAV file, 28 kB)
The latitudinal adjustments involve changes in the cross-sectional area at a given location in the vocal tract. These changes strongly influence the transmission characteristics of the vocal tract (section 24), and have recognisable correlates in voice quality (e.g. "smiling" voice).
( Open jaw , WAV file, 13 kB)
The velopharyngeal settings describe the position of the velum which in its neutral setting facilitates nasality only on certain segments . The other (open or closed) position of the velum during the production of speech segments is regarded as the specific (nasalized or denasalized) setting. The coupling of the nasal resonator dramatically changes the spectral characteristics of the vocal tract. Additional formants and antiformants are produced while the high frequency components are suppressed.
( Pharyngalized voice , WAV file, 21 kB)
( Nasalized voice , WAV file, 18 kB)
According to Laver (1980, 1991) the muscular settings can give rise to two types of voice: tense voce is created by increased muscular tension, whereas lax voice is caused by an overall decrease in muscular tension (compared to the neutral setting). Muscular tension influences all aspects of speech production. In respect to the voice source tense voice (as described in section 7.5) is loud, high-pitched, with harsh phonation and higher subglottal pressure. Lax voice on the other hand is soft, with lower pitch, low subglottal pressure and breathy phonation.
Voice quality is sometimes described using the category of a register of voice. However, this categorization of voices is often misleading. Aregister is defined by Titze (1994:253) as a perceptual category that divides the regions of vocal quality into distinct regions that are maintained over some ranges of pitch and loudness. The register is a categorical feature which is quantally perceived and which is related to categorical perception. Hollien (1974) introduced different registers for speaking and singing. He distinguished between pulse, modal and falsetto registers for speech, but he used chest, head and falsetto registers for singing. These labels are often regarded as the counterparts of the phonatory settings, although the register categories are more complex and involve supralaryngeal settings. Previously, the voice register was mainly seen as the consequence of pitch alone (i.e. as the perceived fundamental frequency of voice). Nowadays, it is seen as a category of vocal fold vibrations (Zemlin, 1988).
8. for a detailed description the reader is referred to Laver (1991).