Below is a schematic representation of the movements of the articulators in the production of a stop. The two articulators for [p,b], for instance, are the upper lip and the lower lip; for [t,d] they are the tip plus rims of the tongue and the alveolar ridge; and for [k,g], they are the back of the tongue and the soft palate. The figure represents the three stages in the articulation of a stop:
During the closing stage (1), the articulators are brought together to form a complete closure at some point in the speech tract, behind which the air from the lungs is trapped.
This is followed by the compression stage (2), during which the air pressure behind the closure will increase as more air continues to be expelled from the lungs.
The third stage is the release of the closure (3): if the articulators are separated abruptly, the air will rush out from behind the closure, which gives rise to the characteristic popping sound of a stop.
Stops that are pronounced with a sudden release are also called plosives, because the sudden release causes the air to literally explode outward. Try and become aware of the gesture represented in Figure 2 by silently mouthing [apa], [ada] etc.
A stop is fully voiced when the vocal cords vibrate throughout its articulation, from the beginning of stage 1 to the end of stage 3, as for /b/ in GA and AN label.
When there is no vocal cord vibration at any time during the articulation of a stop, it is completely voiceless, like /t/ in GA and AN stem.
When vocal cord vibration occurs during part of the articulation of a stop only, the stop is partly devoiced, as in GA bead.
If the stop is voiceless, and the open glottis is maintained for a while during the production of a following vowel or sonorant (that is, stage 3 is also voiceless), the stop is aspirated. We will see that some of the major differences between GA and AN stops are due to a different timing of the beginning or the end of vocal cord vibration relative to the movements of the articulators.