Study of a System of Minimal Speech‐Reproducing Units for Italian Speech

A language maybe synthesized by combining, in a suitable way, elementary speech segments that have been extracted from the utterances of a native speaker of that language. The main criteria for the choice of these elements are discussed. The discussion leads to the definition of a new set of physically determinate linguistic elements that can be obtained by experiment. A preliminary list is reported for the Italian language. From experimental tests, it appears that it is possible to synthesize Italian speech with a small number of elements. These elements may form the basis for research in speech recognition.