The voicing effect in English and Arabic

Abstract EN

Vowels, as many languages exhibit, tend to be longer whenpreceding a voiced consonant than when preceding a voiceless one.

The present study examines the degree of variability in vowel duration induced by the voicing feature of the following consonant in two different languages : English and Arabic to test Chen's claim (1970) that the voicing effect of consonants on vowel duration should be regarded as a strong phonetic universal.

It has been found that in Modem Standard Arabic (MSA) there is only an overall minimal evidence for vowel lengthening before voiced obstruents (6-23 ms at different reading rates).

This insignificant amount of increase in vowel duration does not give support to the above claim but rather indicates that there is a tendency for relatively longer vowels before voiced consonants and thus proves the observation expressed by Lehiste (1970) and Ladefoged and Maddieson (1996) that the amount of increase in vowel duration due to the voicing effect varies from one language to another.

Phonetic voicing in MSA is primary and efficient enough to distinguish homorganic sets of obstruents, and therefore there is no need to replace the voicing feature of consonants by a secondary one as the duration of the preceding vowel.

In other languages where the maintenance of vocal fold vibrations is not that essential to the production of voiced consonants as in English (Lisker, 1957), lengthening of vowels preceding voiced consonants serves as a major distinguisher for the voicing contrast in the phonological system (Raphael, 19'72).

This shows the various patterns of temporal implementation across languages.