RT at bDLF.

The interest of using reaction time measurements was two-fold. Firstly, it enabled the possible behavioural consequences of a local improvement in DLF to be further investigated. This improvement is thought to be linked to both cortical (Robertson and Irvine, 1989; Rajan et al., 1993) and sub-cortical (Harrison et al., 1998) over-representation of the cut-off frequency of a steeply sloping hearing loss. This over-representation has further been shown in animals to modify the salience of lesion-edge frequency tones, and may thus modify behaviour (Willott et al., 1994). We consequently investigated whether over-representation at the bDLF could also alter reaction time.

Secondly, one of the goals of this study was to shed light on the processing of a simple reaction time task. A simple reaction time procedure involves several mechanisms. The first is the encoding of the signal at a peripheral level and its transduction. Next, there is the time of transduction of the signal to the brain, and lastly the time it takes to activate the muscles to press the button. There are many arguments to show a peripheral influence on RT measurement (Chocholle 1940), but the central counterpart is less well-known. If the central component has an influence on RT, then RT would be likely to be altered at the bDLF.

Reaction time did not vary with tone frequency at either the first or the second time of testing. This result was expected, since the tones, although of different frequencies, were of equal loudness. Moreover, there was no change in RT at the bDLF. As stated in the Introduction, this fact argues in favour of a peripheral origin to the relationship between RT and frequency. Another possibility is that the plasticity represented by the frequency discrimination improvement may be too subtle a sign with regard to other mechanisms that may be involved. Cortical factors being just one part of a complex process, cortical plasticity may not be sufficient in itself to induce an increase or decrease in a gross behavioural parameter.