Current linguistic theory rigidly compartmentalizes the "cognitive," linguistic aspects of human communication and the presumed "emotive," paralinguistic elements that occur in both human and nonhuman communication. The segmental phonetic units of human speech, according to this view, are supposed to convey linguistically relevant information, e.g., the vowel distinction that differentiates the English words "bit" and "bet." Emotive, paralinguistic qualities are supposedly transmitted only by means of prosodic modifications like fundmental frequency, amplitude, and tempo as well gestures and facial expressions. Animals, according to this view, make use only of these "paralinguistic" parameters. This distinction is false. The same phonetic feature space is used for both paralinguistic and linguistic communication, and the semantic boundary line between these two aspects of human communication is not sharp. The foundations of human language can be seen in the paralinguistic aspects of human communication and in the vocal and gestural aspects of the communications of other animals. (Author)