A 100-item test for the identification of phonemes was presented to two groups of listeners. One group consisted of 35 patients with normal hearing up to 2000 Hz accompanied by a high-tone loss. The other group consisted of persons with normal hearing who heard the speech stimuli presented through a low-pass filter with a cutoff at 2040 Hz. There were no significant differences between the two groups in overall scores, in probabilities of error for individual phonemes, or in the kinds of error substitutions made. Findings were as follows: (1) the overall scores indicated difficulty in phonemic identification; (2) the phonemes contributing to this difficulty were primarily /b, p, t, k, s, θ/ in both the initial and final positions, /t∫, ∫, f, dƷ, z, v/ in the final position only, and /d/ in the initial position only; and (3) the phonemes substituted for the stimulus phonemes in the initial position were the same in manner of articulation as the stimulus phoneme, and typically only one confusion occurred per stimulus phoneme. The phonemes substituted for the stimulus phonemes in the final position were not necessarily the same in manner of articulation as the stimulus phoneme, and more than one confusion generally occurred for each stimulus phoneme.
Read full abstract