EVENT: Once, twice or thrice upon a time: Audio-visual temporal recalibration is driven by decisional processes

Derek Arnold’s research focuses on issues that arise due to the functional architecture of human sensory processing. Sensory analyses are often initially independent, can take place in different brain regions and be completed at different rates. How then do they combine across space and time in the brain to create subjectively uniform experiences? How do we discern when one event happened relative to another?

Dr. Derek Arnold completed his PhD at Macquarie University in 2003. He then took up an Anglo-Australian Research Fellowship, funded by the Royal Society. In 2006 he took up a continuing position at the University of Queensland, initially as an ARC funded Australian Postdoctoral Fellow, then as an ARC funded Australian Research Fellow, and in 2014 he began an ARC funded Future Research Fellowship.

Title: Once, twice or thrice upon a time: Audio-visual temporal recalibration is driven by decisional processes

“Time is an illusion. Lunch time doubly so” (Adams, 1979). A malleable sense of time is not unique to the bacchanal, but commonplace, resulting in some striking, and some subtle, illusions. For instance, after exposure to an audio-visual asynchrony, a subsequent similar audio-visual test can seem more synchronous than it would without the pre-exposure – an audio-visual temporal recalibration (TR). The underlying cause(s) is unclear. One suggestion is that TR reflects changes in processing speed. This seems unlikely as collaborators and I have shown that one can induce simultaneous opposite TRs simultaneously, and that these tend to be tied to actor identity rather than to spatial location. Another suggestion is that TR results from multiple channels tuned to different temporal offsets, with exposure to asynchrony causing systematic changes in channel responsiveness and encoded timing. A third explanation is that TR primarily reflects changes in decisional criteria. Consistent with this last suggestion, we have found that TR is strongly influenced by task demands. TR is approximately halved by asking people if sounds preceded or lagged visual events, or if timing was indeterminable, as opposed to simply asking if signals were synchronous or asynchronous. We believe the former task encourages participants to adopt more rigorous and stable criteria, highlighting the importance of these processes over changes in the responsiveness of hypothetical channels, for which no firm evidence exists.