Marked results have been achieved in recent years in text-to-speech synthesis. But though intonation of discourse has receivedo considerable attention in the past decade, prosody being one of the main factors, determining the quality of synthetic speech, surprisingly little research has been carried out into synthetic discourse intonation.