Congruent timing in Human, Nightingale, Parotid Wasp and Hawaiian Planthopper vocalisations

Earlier multi-speaker, multi-language measurements found that in purposeful fluent speech, total voice duration was half that of the whole - a maximum Shannon entropy criterion for two state signals. The present aim is to discover whether other species vocalisations are similarly associated with maximum timing entropy control. Data selection has been based on the use of courtship songs since they are purposeful and fluent. These songs are intended for con-specific females and, at present, often too complex for reliable acoustic signal analysis. In consequence it has been necessary, for this preliminary study, to find courtship songs in which the echemes are well demarcated from silence - this has led to a curtailment of potential signal resources. The vocalisation timings in examples of the courtship songs of Nightingale, Parotid Wasp and Hawaiian Planthopper have been analysed. The resulting measurements are compared with those for voice timing in purposeful fluent, human speech. There are three primary findings: O_LIThe same Shannon timing entropy maximising rule is intrinsic to all four vocalisations. C_LIO_LIGaussian distributions of voice/vocalisation timings are present for each of the four signal types. C_LIO_LIIn each case, maximum entropy of voice/vocalisation timing tends to be maintained continually throughout the production of the whole signal. C_LI