English of in L1 and L2 speakers’ read and spontaneous speech

Abstract

This paper compares realizations of the English function word of in read and spontaneous speech produced by Czech and Norwegian speakers with native productions. Acoustic analysis involved word and segment durations, voicing, formant and band energy measurements. Non-natives appeared to produce longer durations than natives but read productions were longer than spontaneous tokens for natives and non-natives alike. Relative segment durations, however, varied between speaker groups. F1-F0 values in of did not differ between groups. Higher F3-F2 values indicated less fronted vowel quality for the non-natives than for the natives. In this case there was no effect of speaking style. Amount of fricative voicing and friction intensity varied systematically with absence or presence of voicing in the following segment.