How Speech Is Produced

Three Systems in Series

Speech production runs through three stages. Almost every phonetic category describes one of them.

1. The airstream

Nearly all speech uses a pulmonic egressive airstream. This is air pushed out of the lungs. Other airstreams exist:

• ejectives: use air compressed by the larynx
• implosives: draw air inward
• clicks: use suction between two closures in the mouth

2. Phonation: the larynx

Air passes through the vocal folds:

• Held close together, they vibrate to make voiced sounds.
• Held apart, air passes freely to make voiceless sounds.

The vibration rate is the fundamental frequency ($F_0$). A listener hears this as pitch:

Larger folds vibrate more slowly. This explains why pitch drops during male adolescence.

3. Articulation: the vocal tract

Above the larynx, air passes through a tube about 17 cm long in an adult male. Articulators change its shape:

• active articulators move (such as the tongue and lower lip).
• passive articulators stay still (such as the teeth and hard palate).

Two further controls shape speech:

• The velum lowers to open the nasal cavity. This makes a sound nasal rather than oral, separating /b/ from /m/.
• The lips can round. This lengthens the vocal tract and lowers resonant frequencies.

Speakers produce ten to fifteen speech sounds per second. Articulators overlap constantly during speech.

Common pitfall: treating voicing as an all-or-nothing property of a sound. It is a state of the larynx over time. Initial English "voiced" stops often lack vocal fold vibration during closure.