Experiment. Critical review of the Emotion in Speech Project in Reading/Leeds. Statistically analysed the prosodic/paralinguistic Roach2000 and emotional GreasleySherrardWaterman2000 annotations of the speech material. Nearly no correlations could be found. This is tentatively explained by the too fine-grained level of detail in both types of annotation, while not controlling for non-emotional factors in the spontaneous speech corpus.