Full Field Masking Causes Reversals in Perceived Event Order

Samson Chota, Douglas McLelland, Louisa Lavergne, Eckart Zimmermann, Patrick Cavanagh, Rufin VanRullenFrontiers in Neuroscience 2020 · DOI

TL;DR

This study demonstrates that a visual illusion, where the perceived order of two brief events is reversed, can be induced by a full-field visual mask, similar to how it occurs during eye movements (saccades). This finding suggests that the mechanisms underlying visual stability during disruptions might be more general than previously thought, not solely tied to saccadic eye movements.

What Problem Does This Paper Address?

Our visual system is remarkably adept at maintaining a stable perception of the world despite constant disruptions, such as our own eye movements (saccades) or sudden changes in the visual input (like flashes or occlusions). Researchers have extensively studied how the brain handles these disruptions, particularly during saccades, identifying phenomena like "saccadic suppression of displacement" and temporal/spatial compression. While some of these peri-saccadic distortions have been shown to occur even without saccades, when the visual input is disrupted by masking, it was unclear if all peri-saccadic illusions could be replicated. This paper specifically investigates whether the reversal of perceived temporal order, a known peri-saccadic illusion, can also be induced by a simple visual mask.

Methods

The researchers employed a temporal order judgment (TOJ) task. Participants were presented with two brief visual stimuli: a red "target" and a yellow "probe," appearing at slightly different times. Crucially, a full-field, pixelated white noise mask was then presented at various timings relative to the target and probe. Participants' task was to report which stimulus (red or yellow) they perceived as appearing first. The key manipulation was varying the stimulus onset asynchrony (SOA) between the target and probe, and the stimulus onset asynchrony between the target and the mask. A control condition without any mask was also included for comparison. To analyze the data, they used maximum likelihood estimates for individual performance and averaged these for group results, employing statistical tests to determine significance.

Datasets

Not discussed in the paper. The study involved human participants performing a controlled experimental task.

Key Insights and Ideas

  • Masking as a Saccade Mimic: The most striking finding is that a full-field mask, in the absence of eye movements, can induce a temporal reversal illusion, mirroring effects previously observed around saccades. This strengthens the idea that the brain's mechanisms for maintaining visual continuity during disruptions are not exclusively tied to the motor act of saccading.
  • Temporal Reversals are Subjective and Confident: Participants didn't just guess; they often reported these temporal reversals with high subjective confidence. This indicates it's not a simple failure to detect one of the stimuli, but a genuine perceptual misordering.
  • Timing is Crucial: The illusion was most pronounced when the mask appeared shortly after the probe, suggesting that the disruption's timing relative to the events being judged is a critical factor.
  • Individual Variability: The study highlights significant inter-individual differences in susceptibility to this illusion, with some participants experiencing it frequently and others rarely, even under identical conditions. This points to individual differences in perceptual processing or attentional mechanisms.
  • Beyond Simple Temporal Compression: While temporal compression is a known peri-saccadic effect, the authors argue that their findings, particularly with longer stimulus intervals, suggest that temporal reversal might involve more than just a simple broadening of temporal perception.

Key Results

  • Approximately 25% of trials resulted in perceived temporal reversals at the shortest stimulus-to-mask intervals.
  • The temporal reversal illusion was observed across various mask onset timings (50, 100, and 150 ms after target onset).
  • The peak effect for reversals occurred when the probe appeared shortly before or overlapped with the mask.
  • Even when participants reported high confidence in their perception, temporal reversals were still observed, suggesting a robust illusion.
  • Individual differences were substantial, with some participants showing strong effects in all masking conditions, while others showed minimal or no effect.

Limitations and Future Work

  • Common Underlying Mechanisms: While the study shows similar phenomena, it acknowledges that it cannot definitively prove that the exact same underlying neural mechanisms are responsible for both saccade- and mask-induced temporal reversals.
  • Confounds: The authors address potential confounds like unintentional eye movements and attentional biases ("prior entry"), arguing against them based on control conditions and the nature of the mask. However, the possibility of subtle interactions remains.
  • Model Development: The paper proposes that a serial updating process, potentially linked to neural oscillations (theta/alpha bands), might underlie these temporal reversals. This offers a theoretical framework for future research.
  • Exploring Other Peri-Saccadic Illusions: The success in replicating temporal reversal with masking suggests that other peri-saccadic phenomena could also be investigated using similar masking paradigms, potentially offering greater experimental control.
  • Further Investigation of Individual Differences: Understanding the basis of the observed inter-individual variability could provide deeper insights into the neural substrates of temporal perception and visual stability.

Summary updated

Abstract

We generally experience a stable visual world in spite of regular disruptions caused by our own movements (saccades, blinks) or by the visual input itself (flashes, occlusions). In trying to understand the mechanisms responsible for this stability, saccades have been particularly well-studied, and a number of peri-saccadic perceptual distortions (spatial and temporal compression, failure to detect target displacement) have been explored. It has been shown that some of these distortions are not saccade specific, but also arise when the visual input is instead abruptly and briefly masked. Here, we demonstrate that another peri-saccadic distortion, the reversal of the temporal order of a pair of brief events, may also be found with masking. Human participants performed a temporal order judgment task, and the timing of stimuli and mask was varied over trials. Perceptual order was reversed on ~25% of the trials at the shortest stimulus to mask intervals. This was not merely a failure of target detection, since participants often reported these reversals with high subjective confidence. These findings update the constraints on models of stability around disruptions.
BibTeX
@inproceedings{chota2020full,
  author = {Samson Chota and Douglas McLelland and Louisa Lavergne and Eckart Zimmermann and Patrick Cavanagh and Rufin VanRullen},
  title = {{Full Field Masking Causes Reversals in Perceived Event Order}},
  booktitle = {Frontiers in Neuroscience},
  year = {2020},
  doi = {10.3389/fnins.2020.00217},
}
Figure