A scientific measure of our visual imagination suggests it is surprisingly limited

The ThinkerBy Alex Fradera

The human mind has been so successful in transforming the material world that it is easy to forget that it too is subject to its own constraints. From biases in our judgment to the imperfection of our memory, psychology has done useful work mapping out many of these limits, yet when it comes to the human imagination, most of us still like to see it as something boundless. But new research in the journal Cognition, on the capacity of our visual imagination, suggests that we soon hit its limits.

Rebecca Keogh and Joel Pearson from the University of New South Wales conducted a series of tightly controlled experiments involving 72 participants, mostly students from the university. In every experiment, participants donned special glasses and sat before a computer screen to view dozens or hundreds of trials each composed of two phases.

In the first phase, the screen was lit up with dotted outlines marking out “imagination spaces”. The participants had to mentally “fill-in” between one and seven of these spaces with a simple pattern, either red with horizontal stripes or green with vertical stripes (similar to those pictured below). Cues in the dotted outlines indicated which pattern to visualize in which space. Participants attempted to keep the imagination spaces appropriately coloured in their minds for six seconds.

Screen Shot 2017-03-29 at 09.43.15

Immediately afterwards, the second phase of each trial determined how successfully the participants had pictured the patterns, and this is where the glasses came into play. They beamed the red pattern into one eye, and the green into the other. When faced with two competing images presented to each eye in this way, the human mind quickly settles on just one image, banishing the other from conscious awareness in a well-known phenomenon called “binocular rivalry”. All things being equal, the battle between competing images is random: half the time one will dominate our perception, half the time the other.

But crucially, the outcome of this process can be influenced by the content of our imagination, especially when the real and imagined share the same location. The researchers took advantage of this and, each trial, always targeted the real images at one of the imagination spaces that had been triggered in the first phase, to see whether the mental pattern in that location was strong enough to tip the binocular battle one way or the other.

When participants only had to fill one imagination space, and the two rival real patterns were beamed to this location, the mental image there strongly influenced the rivalry phase: participants were swayed towards seeing whichever real pattern corresponded with the imagined pattern seventy per cent of the time (fifty per cent would mean no influence).

But once participants were tasked with filling two imagination spaces, the biasing effect of beaming the real patterns onto either one of those imagined patterns dropped to sixty per cent. Imagining three or more patterns led to a further drop-off in the influence of imagined patterns on the process of binocular rivalry. In fact, the experimenters quickly adjusted their design to avoid trials with six or more imagination spaces as it was clear participants couldn’t handle it.

Doubling the imagination phase duration to 12 seconds didn’t improve success for multiple imagination spaces, ruling out the possibility that more imagining just takes more time to get going. Further experiments revealed how the mind deals with greater demands on the imagination: trying to fill more imagination spaces led to rougher mental images at every location, rather than a few filled vividly with others more or less abandoned. This tallies with participants’ descriptions: they said vividness of the imagined patterns suffered as demand increased. This quantity-quality trade-off suggests some kind of resolution issue: think of limited pixels on a screen, or the space available on an artist’s canvas. These two-dimensional examples may be more than an analogy, as visual brain areas have limited anatomical space to represent information, and often do so across two-dimensional maps.

Keogh and Pearson said their results are beginning to map out the “severe capacity limits” of internal representation, and demonstrate how much the architecture of thought depends on the foundations of perception. This is truly important work in understanding the mind, and how much of the world we can truly grasp. Yet, contrasting the effortful yet arbitrary visualisation investigated in this study with the rich, spontaneously arising visions I’ve personally experienced during meditation – or the overwhelming sense of reality we’ve all experienced in dreams – I’m left wondering how these limits may vary according to our states and motivations. That’s worthy of future research, because feeling out the firmament of the imagination is a most worthwhile goal.

The perceptual and phenomenal capacity of mental imagery

Main image via gettyimages.co.uk

Alex Fradera (@alexfradera) is Contributing Writer at BPS Research Digest

6 thoughts on “A scientific measure of our visual imagination suggests it is surprisingly limited”

  1. Don’t serve images that can induce the McCollough Effect. That’s dangerous and irresponsible, and you should know better.

    1. While staring at the images for 15 minutes or more can cause the McCollough effect to last up to 3 months, it’s not dangerous given that the effect isn’t permanent.

Comments are closed.