Bite Block Speech Under Anesthesia

The following audio recording is from an experiment by Bjorn Lindblom and colleagues around 1970 in which a speaker attempted to produce /i/ with bite blocks of different sizes, with and without surface anesthesia (xylocain) applied to the oral cavity. Note how production is only seriously disrupted when anesthesia is applied and an unnaturally large bite block size is used. According to the DIVA model, when there is a large bite block and no anesthesia, the somatosensory feedback control system sends compensatory movement commands until the tongue configuration for /i/ is reached, thereby compensating for the bite block even before sound is produced by the vocal folds. When anesthesia is applied, however, the talker cannot sense when the tongue is in the right place, impairing the ability of the somatosensory feedback controller to compensate for the bite block. Initially the speaker “over-compensates” by pushing the tongue all the way to the palate, closing off the vocal tract. Only after hearing the consequences of this over-compensation can he correct the production using auditory feedback control mechanisms. [Sound file courtesy of Bjorn Lindblom.]