Articulatory Control in a Deaf

While many children who are born severely or profoundly deaf, or become deaf in infancy achieve intelligible speech, the vast majority do not. Speech intelligibility is fairly well correlated with residual hearing (Boothroyd, 1970; Smith, 1972) at least until 90dB, and overall intelligibility is well correlated with the percent of segmental errors, and to a lesser extent with suprasegmental deviancy (Levitt, Smith, & Stromberg, 1974). While many educators of the deaf would claim that the characteristic unintelligibility of deaf speakers is a consequence of faul ty teaching practices (Haycock, 1933; Ling, 1976), independent investigations have been remarkably consistent in showing similar patterns of segmental and suprasegmental errors in the speech of deaf talkers trained in a wide variety of programs (Hudgins & Numbers, 1 Smith, 1972; Levitt, Stark, McGarr, Carp, , Barry, Vele z, Osberger, Leiter, & Freeman, Note 1; Johnson, 1975). Furthermore, experienced teachers of the deaf can discriminate between deaf and non-deaf speakers from disyllables produced by both groups (Calvert, 1961), and experienced listeners of the deaf are better than naive listeners in decoding deaf utterances (McGarr, 1978). If we accept the point of view that there is a "deaf speech" 1 pattern, not dependent at least on the fine-grained details of the training procedure, we may ask what are its characteristics? do the deaf sound as they do? villy are they unintell ible?