Page 68 - It' about time: Studying the Encoding of Duration

The duration after-effect does not reflect adaptation to perceived duration Introduction Recently, it has been proposed that duration-tuned mechanisms underlie the encoding of duration (Heron et al., 2012; Ivry, 1996). According to this idea, the brain contains groups of duration-tuned neurons that respond selectively to specific ranges of durations. Summation of the population response of these groups of duration-tuned bandpass-neurons (or channels) allows for implicit temporal signals to be transformed into an explicit code for duration that is both accurate and reliable. This explicit signal can then be stored, manipulated, and used to guide subsequent behavior (Heron et al., 2012). The proposal for channel-based encoding of duration is very similar to the mechanisms that are thought to underlie the encoding of a range of other sensory features such as orientation (Gibson, 1933; Hubel & Wiesel, 1959), motion direction (Albright, 1984; Anstis et al., 1998), pitch (Romani et al., 1982), and numerosity (Burr & Ross, 2008; Harvey et al., 2013). Similar to these other stimulus features, support for the channel-based encoding of duration comes from studies demonstrating duration tuning in both behavioral and neuronal 4 responses (Hayashi et al., 2015; Heron et al., 2012). For example, several studies have shown that adaptation to duration leads to a duration after-effect (DAE) for subsequently presented durations (Heron et al., 2012, 2013; Li, Yuan, & Huang, 2015; Maarseveen et al., 2017; Maarseveen, Hogendoorn, Verstraten, & Paffen, 2018; Shima et al., 2016). More specifically, these studies demonstrate that adaptation to a particular duration in one modality causes the perceived duration of subsequent durations in that same modality to shift away from the adapted duration. For example, after adapting to a visual event lasting 400 ms, the perceived duration of visual events with a shorter duration (i.e. 200 ms) will decrease, while the perceived duration of events with a longer duration (i.e. 800 ms) will increase. These results are taken to reflect selective adaptation of duration-tuned neurons, resulting in a shifted population response for durations close to the adapted duration (Heron et al., 2012; Kohn, 2007). Similarly, studies have shown that training observers to discriminate durations leads to increased performance on the trained but not the untrained durations (Bueti & Buonomano, 2014). In line with the adaptation results, these results suggest channel-specific training benefits and are similar to the results observed for discrimination training in other features encoded in a channel- based fashion (Bueti & Buonomano, 2014; Schoups, Vogels, Qian, & Orban, 67