Synthesizing Different Speech Styles

The following is concerned with the linguistic components of expressive speech synthesis. It tries to fulfil the role that Huckvale (2002) assigns to the linguist in speech science for speech synthesis. Linguists find models and parameters for the description of language. The semiotic foundations for the study of expressive herein have to be modelled upon a concept that every sign has got a content, a structure and a rendering. This view is expanded to the degree that content is not simple but comprises an information layer, a paralinguistic content layer and an extralinguistic content layer. These three layers need to be examined closely. As prosody is assumed to be the main carrier for extralinguistic and paralinguistic content in speech these types of content need to be related to the prosodic structure in speech and the rendering of prosody in speech synthesis.