Perception of certain phonetic variables∗

Abstract

R attempts have been made -to utilize speech reception hearing tests to indicate hearing loss for pure tones and to reveal medical diagnoses, uses distinct from measuring hearing loss for speech. The common rationale for such tests has been that speech sounds have characteristic frequency regions and that distinctions among speech sounds are dependent upon hearing acuity for those regions. However, these tests generally have not been useful except for rather gross types of aural pathology or audiometric hearing loss. Other research has suggested that factors such as the influence of sounds upon each other or the position of a sound in a syllable may be important to discrimination among speech sounds. With the development of the sound spectrograph an objective record of the influence of one sound upon another and of the transitional sounds produced is possible. Joos suggested that cues regarding consonant voicing, i.e., whether or not a consonant was produced with vocal tone, and pressure pattern, i.e., whether a consonant were a plosive, a continuant, or an affricate, were the only information a listener gets from the consonant heard in context; other information for identifying the consonant comes from its in-