Voicelessness and palatality

Palatal approximation in English is generally accompanied by vocal fold vibration: [j] is normally voiced. However, it can also be accompanied by voicelessness in clusters with voiceless consonants. For instance, in ‘beauty’ it is voiced, but in ‘pewter’ it is voiceless. When voiceless, there is greater airflow across the glottis because it is open. A greater quantity of air means that the pressure behind a constriction can build up more quickly; and in turn, this means that it is easier to generate noisy turbulent airflow with voiceless sounds than with voiced sounds. So strictly speaking, most of the sounds that result from the combination of voicelessness + palatality + open approximation are in fact not approximants but fricatives.

There are several possible ways to transcribe the combination voicelessness + palatality + friction. An allophonic transcription could modify [j] by marking voicelessness using the diacritic  or . Words like ‘pewter’, ‘few’, ‘cue’ and so on can be transcribed with . This captures the structural similarity between the voiced and the voiceless pairs, so for instance for ‘beauty’ we would have [bju-] and for ‘pewter’ we would have , and it makes it obvious in a visual way that there is a relationship between the sequences in both cases.

More narrowly and impressionistically, we could use the symbol for a voiceless palatal fricative, [ç]. This makes it more obvious that as well as palatality there is voicelessness and friction; in the transcription , friction is more implicit than explicit.