On the acoustic theory of coarticulation and reduction

The paper offers a linear model of coarticulation and reduction on the acoustical level. Simple equations are derived for calculating formant frequencies with regard to the phoneme environment and tempo of speech. In speech synthesis these equations used for calculating formant frequencies taking into account the effects of reduction and coarticulation. In speech recognition they are used for calculating the decoarticulated and de-reduced values of formant frequencies in the speech flow.