Abstract: | This paper tests the hypothesis Licensing by Cue (Steriade 1997) applying it to the distribution of Russian plain-palatalized contrast in coronal stops /t/ vs. /tj/ in two environments (V_C, V_#). The hypothesis holds that the maintenance of the contrast should correspond to more acoustic information in the signal and higher identification rate, and that its neutralization should be accompanied by fewer cues and lower recognition of the segments. The results of acoustic and perceptual experiments do not fully support the hypothesis: while the relative acoustic and perceptual salience of the contrast before three consonants (_#k, _#n, _#s) correlates with within-word neutralization patterns, the lack of neutralization after three vowels (a_, u_, and i_) does not follow from the acoustic and perceptual results. The findings suggest that acoustic and perceptual factors play a certain role in maintenance and neutralization of phonological contrasts, however, the mapping between acoustics and phonology is not direct. |