Martha W. Burton
415 Moore Bldg., Dept. of Psychol., Penn State Univ., University Park, PA 16802
Karen E. Robblee
Penn State Univ., University Park, PA 16802
In Russian, voicing assimilation is usually regressive with the voicing of obstruent clusters conditioned by the final obstruent. However, there is conflicting phonetic evidence from other languages such as Polish on whether neutralization of voicing is complete [cf. W. Jassem and L. Richter, J. Phon. 17, 317--325 (1989); L. Slowiaczek and D. Dinnsen, J. Phon. 13, 325--341 (1985)]. This study investigated the effects of voicing assimilation across prepositional boundaries, an environment where voicing assimilation is believed to occur in Russian. Three native speakers of Russian produced sentences containing prepositions ending in a voiceless stop or fricative. These prepositions were followed by word tokens with initial obstruents that were voiced or voiceless. Thus there were four stop-fricative combinations [ts,dx,tz,dz] and four fricative-stop combinations [st,sd,zt,zd]. Acoustic analyses of these tokens included duration measurements and spectral analyses. Preliminary results indicate that a pattern generally consistent with voicing assimilation emerges. However, for some speakers there was variation depending on the particular combination of fricative and stop. The findings of the study will be considered in terms of phonetic and phonological theories of speech.