On the psychological reality of the phoneme: Perception, identification, and consciousness

each list for the presence of either a specified word or an initial syllable of a word. Reaction times to respond to these targets were recorded. Reaction times were significantly shorter when the target was a two-syllable word. In Experiment II, 45 subjects listened to the lists and monitored for either a two-syllable word, the initial syllable of a word, or the initial phoneme of a word. Reaction times were shortest for word targets and longest for phoneme targets. Implications of these findings for the "reality " of the units was discussed. A distinction between perception and identification was introduced and, in addition, a proposal about a determinant of consciousness was forwarded. One of the major problems to be faced by any model of sentence comprehension results from the fact that the acoustic signal is not reliably segmented into "language-sized" units. The work of Liberman and his asso-ciates (e.g., Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967) has amply demon-strated that there is rarely a one-to-one correspondence between the acoustic signal and the units of perception. The same signal will be perceived differently in different linguistic environments and, in some cases, quite different signals will be perceived as being identical. In addition, all the important acoustic cues for some perceptual entity may not occur at the seemingly proper or expected place in the speech stream. In some cases, for example, a critical cue for a consonant is given by the relative length of the preceding vowel (Chomsky & Miller, 1963). These facts indi-cate that the speech decoding processes cannot be simple ones. The present paper is primarily concerned with characterizing the units of 1 These studies were supported by NSF Grant GS-3285 to the University of Texas at Austin. The authors than