Prosody is “the rhythmic and intonational aspect of language” and it can be a useful feature in language recognition and absolutely critical when the language is tonal like Chinese. Researchers are developing prosody detection methods with the intent of deciphering emotions. Determining the models of various speech characteristics is a first step, and sometimes a daunting task as the various characteristics of volume, pitch, and pauses must be identified across words and sentences. Once the patterns are established, the recognizer can make the matches and guess that a speaker is angry because they are speaking louder and with exaggerated emphasis. Of course if yelling at a voicemail system gets you quickly transferred to a live person, won’t everyone do it? Perhaps the next step will be judging the sincerity of the angry tone.