Sonority

Before we start using the concept, we need to define what sonority is. This, in itself, is not an easy task, as it is also far from being uncontroversial. For pedagogical purposes, we will keep it as straightforward as possible. The sonority of a sound is primarily related to the degree of opening of the vocal tract during its articulation. The more open the vocal tract is for a sound, the higher its sonority will be. Thus vowels, which are produced with a greater degree of opening, will be higher on the sonority scale than fricatives or stops, which are produced either with a narrow opening or with a complete closure of the articulators. The second, and relatively secondary (ancillary), dimension is the sound’s propensity for voicing. This becomes relevant when the stricture (degree of opening) is the same for two given sounds; the sound that has voicing (e.g. voiced fricative) will have a higher degree of sonority than its voiceless counterpart (e.g. voiceless fricative). Putting all these together, we can say that low vowels (/æ, ɑ/), which have the maximum degree of opening, will have the highest sonority; and voiceless stops, which have no opening and no voicing, will have the lowest sonority. The remaining sounds will be in between.

One finds different hierarchies of sonority in different books and manuals. However, the differences are in details rather than the basic relative ordering. We adopt the following 10-point scale suggested by Hogg and McCully (1987):

Having stated the relative sonority of sounds, we are now ready to look at the number of syllables in words as peaks of sonority. In auditory terms, the sonority peak is more prominent than the surrounding segments. Since vowels and diphthongs are higher in sonority than other segments, they typically occupy the peak positions in syllables. We show this with the following displays for publicity and condemnation:

The principle of peaks of sonority correctly identifies the number of syllables, four, in these two cases.

As we saw earlier, in English we can have syllables that do not contain a vowel. In these cases, the most sonorant consonant will be the syllable peak (i.e. syllabic consonant):

Since the existence of syllabic consonants is due to the deletion of the reduced vowel [ə], they are confined to unstressed syllables. In stressed syllables, we always have full vowels that will assume the syllabic peaks; this leaves no chance for the consonant to be syllabic.

Although the principle of equating the sonority peaks to the number of syllables would hold for thousands of English words, it does not mean that it is without exceptions. We must acknowledge the fact that some English onset clusters with /s/ as the first consonant (e.g. stop [stɑp]), and coda clusters with /s/ as the last consonant (e.g. box [bɑks]), do violate this principle.