In this chapter we will discuss the basic aspects of the source-filter theory of speech production that we introduced in Chapter 2. In the chapters that follow we will consider the acoustic characteristics that differentiate many of the sounds of speech and the articulatory gestures that underlie speech. The source–filter theory makes it possible to relate the acoustic and articulatory levels in a quantitative and coherent manner.

The laryngeal source

Let us start by considering the production of a voiced sound, i.e. one in which the vocal cords interrupt the air flow from the lungs, producing phonation. The larynx can serve as a source of sound energy. It can produce a periodic wave through the process of phonation, as for example in the word mama. It can also generate a “noise” source, e.g. the source during the [h] sound of Hop! The source of sound energy when we whisper is the noise generated through air turbulence at the glottal constriction. The vocal cords do not have to move during a whisper. Many of the sounds of speech involve sources of sound that are generated through turbulent air flow through constrictions at other parts of the airways of the vocal tract. However, it will be useful to begin our discussion of the source–filter theory of speech production by considering voiced sounds in which the larynx produces a periodic source.