The phase of pre-stimulus alpha oscillations influences the visual perception of stimulus timing

Over the last decade or so, there’s been an absolute explosion in the interest in neural oscillations and their role in perception. In particular, I and others are very interested in how neural phase, assumed to reflect fluctuations in neuronal excitability, affects perception on a moment-to-moment basis. A new paper by Alex Milton and Christopher Pleydell-Pearce uses EEG to examine the role of neural alpha phase (in the 8–13 Hz range) in the perception of timing – in this case, asynchrony versus simultaneity of visual onsets.

Participants are cued (validly or invalidly) to either the left or right, and then two peripheral LEDs are illuminated with a stimulus onset asynchrony (SOA) chosen to keep asynchrony detection at threshold for an individual participant (sometimes they are simultaneous, but that’s rare and only to estimate false-alarm rates). When the LEDs were illuminated during the trough of the alpha oscillation (measured over a handful of left, posterior sensors), they were more likely to be correctly perceived as asynchronous, but when they were illuminated during the peak of the oscillation, they were more likely to be incorrectly perceived as simultaneous. The results replicate and extend older work by Varela. And they provide a new source of trial-to-trial variability in asynchrony judgments that may also have individual-differences components due to e.g., differences in individual alpha frequency.

I found myself wondering while reading what the explanation for their result was – specifically, was asynchrony more likely to be perceived because the individual events and their onsets were better perceived during more excitable phases of the neural oscillation [based on my current understanding of near-threshold detection/discrimination data]? –OR– are individual LED onsets that are “transmitted” during successive alpha cycles perceived as separate, but bound if they end up in the same alpha cycle together [a la Lisman & Idiart’s theory that individual items in working memory are “stored” in single cycles of a high-frequency gamma oscillation that are nested in a single cycle of a low-frequency theta oscillation, and consistent with Varela’s ideas]? The authors address this question in the Discussion, and assign the two possibilities to the two sides of the debate about whether perception and underlying “processing epochs” are continuous or discrete, respectively. They suggest that fluctuations in neuronal excitability leading to enhancement of the perception of the LEDs and their temporal relation would be an unlikely mechanism to strictly quantize sensory input (and suggest that their own results are more compatible with a continuous view of perception). On the other hand, assuming that the LEDs might be perceived as asynchronous when they ride along in separate alpha cycles is compatible with a “temporal framing” hypothesis; sensory information is gated into discrete “packets” (see for example VanRullen & Koch, 2003).

The arguments for continuity generally appeal to intuition; it’s of course true that our perception of the world flows from one moment to the next without sharp boundaries. But the most basic building block of brain function, a spike from a single neuron, is all or none – a neuron fires or it doesn’t. In between, there are psychophysical and neural data that can be interpreted as supporting both views. So the issue is far from solved. I dislike fractionation and dichotomies, especially in the context of brains, which seem quite hard to parcel up cleanly. So I’m a fan of the idea that both continuity and discreteness are not only present, but necessary for brain function and cognition (Fingelkurts & Fingelkurts, 2006). I don’t have time in a short blog post to talk about HOW perception and cognition might arise from a critical combination of continuous and discrete neural processes, but highly recommend the cited paper as supplementary reading. The authors suggest directions for future research (such as examining symptoms of certain neuropsychological disorders) that may help to better understand the continuity/discreteness trade-off. And with a better understanding of how both discreteness and continuity might be essential for consciousness and cognition, our interpretation of results like those of Milton and Pleydell-Pearce might allow us some insight into neural mechanisms that doesn’t rely on assuming one or the other, but not both, must be true with respect to continuity vs. discreteness.

–Source article: Milton & Pleydell-Pearce.The phase of pre-stimulus alpha oscillations influences the visual perception of stimulus timing. NeuroImage.