On The Perception of Unreleased Voiceless Plosives in English

An experiment concerned with the possibility of distinguishing auditively between unreleased /?/ and unreleased /p/, /t/, and /k/ in English was carried out, with two groups of hearers, one listening by means of earphones, one listening by means of loudspeaker. It was found that correct identification of all the voiceless stops largely depended on the way in which the sounds were (re)produced and conveyed to the hearers. It was also found that correct identification very much depended on the quality of the preceding vowel sound, especially in the case of /p/, /k/, and /?/. Under the best listening conditions /?/ was identified with considerable accuracy. Under less favourable conditions the identification was—in part—scarcely better than pure chance.