Direct Measurements on Animal Preparations

I have been interested in cochlear mechanics my whole career — now for over 34 years. The first half of that time was spent primarily on direct measurements of basilar membrane motion in animal models (guinea pig and briefly in gerbil), while the second half has been directed primarily at various aspects of human data, as obtained through otoacoustic emissions. The questions asked have been essentially the same, and always technically very exacting.

Virtually from the outset my records were dynamic of the kind of dynamic behaviour which was never seen in neural recording, or indeed using the Mössbauer technique because of its averaging time. My recordings showed substantial behaviour which might have been smeared over by signal averaging. I received flak from other investigators who claimed my preparations were damaged, or not in physiological condition and this led to increasing efforts to resolve the issues raised, with better surgical technique, better controls on running the experiments. There is no doubt in my mind in hindsight that my experimental techniques with the approaches used (now of course considerably dated) were second to none at the time, with no effort spared. The first direct measurement technique used a capacitive probe. This technique required draining the basal turn of the cochlea for the period of the measurement. Using this approach provided the first confirmation of the nonlinear compression seen by Bill Rhode. It was a decent achievement, not primarily because of the confirmation, but because it showed why other had failed to see it. Their preparations were not in good condition. I saw it because I showed my preparations were in good condition, at least for long enough to strongly associate the nonlinear behaviour with low hearing thresholds. Indeed my doctoral dissertation from the Auditory Lab at the University of Western Australia gave very strong clues that basilar membrane was as sharply tuned as the neural tuning curves and that refined experiment would show it. At that point I was already in the throes of finalising my dissertation and of relocating to the US (1981), and that job of finally nailing the story fell to Sellick, Patuzzi and Johnstone (1983).

I have always believed my data which suggested that the basilar membrane moved by much larger low-frequency displacements than theory required to account for the sharp tuning. The important aspect of the capacitive probe data was that the static position of the basilar membrane was changing systematically and that this displacement was also frequency dependent and changed polarity. I was absolutely sure that this slow displacement change was important by virtue of its potential as an analog for hair cell dc potentials and as a regulator of outer hair cell activity, but at the time my capacitive probe technique required to drain the cochlea and it was virtually impossible to conduct an experiment and produce believable data. The three years in St.Louis brought a partial clarification, but it was clear that it would be necessary to repeat the experiments using a technique which did not require draining of the cochlea, and risking loss of peak performance. It was supreme irony that I myself, who had first nailed the connection between preparation condition / hearing threshold and nonlinear compression (LePage and Johnstone, 1980) received so much grief from other investigators for the very same reason.

By 1990 I had repeated the experiments and found the expected much larger dc displacements, so large in fact that they were universally treated as artifacts. Virtually nobody was prepared to give me the benefit of the doubt that seeing these large motions might be a sign of good experimental technique. (One still is drawn to suspect that those who have since shown dc components to be vestigial at the basilar membrane level, are not conducting the same experiment which I conducted). The thinking was universally tied to the notions that all motion of the basilar membrane due to cochlear activity must be produced by the outer hair cells, and must be small in keeping with theory.

As a musician I have always been interested in the Helmholtz resonance theory (1885) of the cochlea which was subsequently revived by Gold in 1948. Several capacitive probe recordings from guinea pigs had produced data which were suggestive of resonant behaviour. Important features were the apparent difference in the pattern of basilar membrane movement across the basilar membrane radially. Adjacent to the hair cells (the actuate region) the movement could have been described essentially as a twitch response, while the pectinate zone rang like a resonance structure, almost as if these “strings” had been struck or plucked. Secondly I saw very interesting behaviour at frequencies above the best frequency of the place being measured, such as evidence of two processes working in opposition. I presented these data and ideas at the 1990 Mechanics Meeting in Madison Wisconsin under the title Helmholtz Revisited.