Individual differences in auditory bistable perception: Evidence from ERP and longitudinal behavioral tracking

The Shepard tones are a type of auditory illusion created using multiple octaves of the same musical note in varying loudness. In 1986, Diana Deutsch demonstrated the Tritone Paradox by sequentially playing two Shepard tones separated by 6 semitones, which removed the physical cues of pitch and relative note distance, thereby creating an ambiguous, bistable auditory stimulus that can be heard as either ascending or descending by the participants. Deutsch showed that long-term factors, such as people's native language accent, played a role in resolving the ambiguity, but little work has been done to explore the effect that the immediate auditory environment may have on such perceptual experience. The current study aims to investigate if varying auditory preceding contexts (one tritone pair played before the target tritone pair) can affect the perception of the Tritone Paradox (Experiment 1), thereby shedding light on the criteria and limitations of how the auditory system interprets the acoustic world. Furthermore, past research has shown a large between-individual differences in how the Tritone Paradox is perceived, but no study has explored within-individual variations across time as most studies have simply assumed that there is no variation in people’s perception of tritones over time. Here we test this assumption with a longitudinal design coupled with event-related potentials (ERP) by tracking each participant’s perceptual reports longitudinally over the course of 9 weeks (Experiment 2). In Experiment 1, we observed no preceding-context effect on the perception of subsequent tritone pair. In Experiment 2, we found different levels of stability amongst different individuals in perceiving the Tritone Paradox, with notable differences between stable-percept and unstable-percept groups in ERP at 200 ms after the tritone. To our knowledge, this is the first study that documents within-individual variability across multiple time points. Together, these experiments show that variabilities in the perception of Tritone Paradox perhaps do not arise from the immediate auditory environment, but from within.