During phonation direct examination of the avian vocal organ, the syrinx, is difficult. Models describing sound generation are therefore based on indirect evidence such as physiological observations, excised syrinx preparations, and theoretical studies. The "classical'' models for songbirds and several non- songbird groups postulate that vibrations of thin membranes (e.g., medial tympaniform membranes - MTM) generate sound. Alternatively a whistle mechanism has been proposed. We studied the syrinx of songbirds and non- songbirds directly by filming it through an endoscope during brain stimulation-induced or spontaneous vocalisations. These endoscopic images imply a different phonatory mechanism. In preparation for phonation, soft tissue masses are moved into and close the syringeal lumen and vibrations of these structures always accompany sound generation. In addition, the presumed sound generating membranes of the "classical'' model were disabled. Birds with disabled MTM not only were able to phonate, but also the vocalisations remained nearly intact, further supporting the new model of sound generation. The low temporal video resolution did not allow a detailed study of tissue vibrations. Using a custom-built optic vibration detector, however, we recorded frequencies at the syringeal source that match those present in simultaneously recorded sounds. This provides strong evidence that vibrating structures generate the vocalisations and thus does not support the whistle hypothesis. (Supported by Danish National Research Foundation, NIH, and NSF)
Abstract: