QnAs with Randolph Blake

Psychologist Randolph Blake has been fascinated for most of his career by a form of perceptual magic called binocular rivalry: Dissimilar images presented simultaneously one to each eye take turns perceptually disappearing and reappearing unpredictably. Furthermore, it is impossible to hold just one of those images in conscious awareness indefinitely. Blake started studying binocular rivalry as a graduate student at Vanderbilt University, and has since become an expert on the topic as he has moved to Northwestern University and back to Vanderbilt, where he is currently the Centennial Professor of Psychology and professor of Ophthalmology and Visual Sciences. Blake is also professor of Brain and Cognitive Sciences at Seoul National University. Blake was elected to the National Academy of Sciences in 2012, and his Inaugural Article attempts to meld two conflicting theories of visual rivalry. Blake spoke to PNAS about his interest in visual rivalry.

Randolph Blake. Image courtesy of Elaine Blake.

PNAS: Why has binocular rivalry fascinated you for so long?

Blake: Because rivalry implies that neural representations of the two conflicting images are competing for dominance within the brain. If we can figure out the details of that competition we will have learned something about the neural bases of visual perception.

PNAS: There has been debate about whether binocular rivalry occurs at low- or high-level processing. Why does it matter?

Blake: The debate matters because the answer dictates where one may find the neural events causing rivalry and, more importantly, the extent to which those neural events may also be involved in other aspects of vision, including perceptual awareness.

PNAS: How did the debate start?

Blake: In 1989 I published a theory of rivalry that was an unabashedly low-level account (1). In a nutshell, I thought of rivalry as a default outcome when binocular fusion and stereoscopic depth perception proved impossible. I still think there’s merit to that idea, but more recent work reveals aspects of rivalry governed by high-level processes within the visual system, including processes mediating attention and global perceptual organization. All of those factors are capable of reinforcing the perceptual dominance of the currently visible rival stimulus. To what extent those factors influence a suppressed stimulus remains arguable.

In addition, Nikos Logothetis and his colleagues discovered what we now call stimulus rivalry, which, like binocular rivalry, is characterized by periods of exclusive perceptual dominance of one stimulus for several seconds at a time when dissimilar images are viewed by the two eyes. However, while with binocular rivalry a given image remains associated with a given eye during the entire viewing period, with stimulus rivalry, the two dissimilar images are rapidly (three times per second) and repetitively swapped back and forth between the two eyes. What’s confounding is that with stimulus rivalry, the images are switching seemingly too fast for you to have a period of two to three seconds of eye dominance. If one eye remained dominant, you should just see a rapid fluctuation between two pictures, but that’s often not what happens. So stimulus rivalry was construed by some as evidence against eye-based competition and in favor of competition between high-level object representations. In 2002, Nikos and I published a widely cited paper (2) proposing a compromise, assigning the two forms of rivalry to different stages of processing within the visual hierarchy, with conventional rivalry having an eye-based component and stimulus rivalry being a higher level, purely binocular phenomenon. Our proposal amicably provided an amalgam of the two theories, but it still left important questions unanswered.

PNAS: Your Inaugural Article takes this compromise a step further; can you explain what you have found?

Blake: For years I remained unsettled by my inability to parsimoniously integrate the two forms of rivalry; so, my colleagues and I set out to discover whether stimulus rivalry and binocular rivalry might be more closely interrelated than thought. We designed two critical experiments, both of which showed that stimulus rivalry, in fact, does have an eye-based component to it and, moreover, that stimulus rivalry strongly interacts with binocular rivalry. The two, in other words, seem at least partially to share common neural processes. Moreover, we developed a neural model that integrates traditional binocular rivalry with stimulus rivalry. We think what’s happening is that dominance is switching rapidly back and forth between the left eye and right eye, following the stimulus. We’re not saying that both forms of rivalry are exclusively low-level phenomena, for, as I said before, so-called “top-down” influences can impact the dynamics of rivalry. However, we do believe that stimulus rivalry and binocular rivalry are more closely related than heretofore believed. We look forward to seeing how these new ideas are received, and we hope our model stimulates additional investigations of these beguiling phenomena.

PNAS: How do you think these findings will be received?

Blake: People in my field seem comfortable with the current truce built on coexistence of low-level and high-level forms of rivalry. Our paper could be construed as a challenge to that truce, but that’s not a fair characterization. In fact, our paper aims at unification, bringing a coherent, integrated account of these apparently different versions of rivalry.