Purchase this article with an account.
SHUCHEN GUAN, XINGNAN ZHAO, YINGZI XIONG, CONG YU; Supramodal representation of subsecond time and the impact of its training on unimodal temporal interval discrimination. Journal of Vision 2020;20(11):1721. doi: https://doi.org/10.1167/jov.20.11.1721.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Whether subsecond timing relies on a centralized clock, or on distributed temporal mechanisms, has been a central theme in time research. Current views favor distributed mechanisms because of the interval specificity in temporal interval discrimination (TID) learning, which contradicts a dedicated centralized clock account.
We first ran a double training procedure to eliminate interval specificity in learning of TID with intervals marked by pairs of auditory beeps. TID learning was initially interval specific: TID learning with a 100-interval after 5 sessions of practice reduced TID thresholds by 42.8±10.0% at a 100-ms interval (p<0.001), but it had no significant impact on TID thresholds at 200-ms (15.9±6.7%, p=0.15). However, when TID training at 100-ms was paired with auditory frequency discrimination (FD) learning at 200-ms in alternating blocks of trials, TID thresholds at 100-ms and 200-ms were both improved (41.8±8.9% & 32.7±1.6%, p<0.001). A control experiment excluded the possibility that TID improvement at 200-ms was caused by FD learning. Similar double training procedures also enabled complete TID learning transfer between visual and auditory modalities. These results suggest an interval-unspecific supramodal representation of subsecond time. Training may refine this centralized clock-like representation, so that learning can transfer across intervals and modalities.
Next we tested whether supramodal TID training could improve unimodal TID performance. Participants practiced supramodal TID of a 200-ms interval defined by an auditory and a visual signal (AV), which not only reduced AV TID thresholds by 36.7±5.4% (p<0.001), but also reduced auditory-auditory (AA) TID at 200-ms by 27.8±4.1% (p<0.001). Continued training of AA TID failed to further improve AA TID thresholds (0.5±9.2%, p=0.96). This finding indicates that precise subsecond timing can be achieved through training of more centralized supramodal timing mechanisms, with no necessity of engaging more peripheral unimodal timing mechanisms.
This PDF is available to Subscribers Only