Our Visual Imagination Is Severely Limited. Here Is Why That Matters.

When we imagine multiple images, we create multiple maps that compete for space in the brain’s visual areas. The implications are wide-ranging.

Imagine you are at Ikea to pick up a sofa for your new flat. You see one you like, a wine-coloured two-seater with big soft cushions. You imagine what it would look like with your current furniture, and decide that’s the sofa you want. As you continue meandering through the store you find a nice industrial-style lamp and coffee table, so you try to imagine what they might look like with the sofa. But imagining all three items together is more difficult than just imagining the sofa alone. How many pieces of furniture do you think you could rearrange in your mind? Is there a limit to how much we can imagine at once, or is our imagination truly unlimited? 


Limitations to our imagery can constrain what we are able to achieve, both in daily life and in therapeutic interventions.

This is the question that my supervisor and I tried to answer in our lab at the University of New South Wales recently. Instead of furniture, we used simple shapes known as Gabor patches, which are essentially circles with lines through them. We also used a visual illusion known as binocular rivalry. Binocular rivalry occurs when two different images are shown, one to each eye, and instead of seeing a mix of the two images you see only one of them, either the image that was presented to the left eye or the image presented to the right eye. Previous work by Joel Pearson (my supervisor) has shown that simply imagining a Gabor patch, or seeing a very weak Gabor patch, will make you more likely to see that image in a subsequent binocular rivalry display.

For example, if I asked you to imagine a red Gabor patch for a few seconds, and then presented you with a binocular-rivalry display comprising of one red and one green Gabor patch, you would be significantly more likely to see the red image rather than the green one. This is known as priming, and is often measured as the percentage of trials primed (that is, the percentage of times that an individual sees the imagined image in the subsequent binocular-rivalry display). Since this task has only ever been studied using one image, we decided to test just how many items we could imagine at once. If we were able to imagine an unlimited number of items, then we should see the same level of priming for one or multiple images.

We started off ambitiously, asking participants to imagine anywhere from one to seven images. We presented them with cues that indicated how many, and what color and orientation Gabor patches they should imagine. Importantly, the cues we gave them were present during the entire time the participants imagined, meaning there was no confounding effect of memory, as they didn’t have to remember which image to imagine where. We found that our participants were limited in what they were able to imagine, with their priming falling to chance levels when they had to hold about three or four items in their mind. We followed up this study with a few more experiments, and found that our subjects also rated their visual imagery as less vivid when they were required to imagine more items, and that the precision of these items decreased when they were required to imagine more than one image at once.

Love Undark? Sign up for our newsletter!

So it can indeed be shown that there are severe limitations to our visual imagination. Why? It is likely that visual imagery-capacity limits arise somewhere in the visual-imagery brain network, which spans frontal and posterior visual areas of the brain. The frontal areas are thought to be responsible for the control and creation of visual imagery through top-down connections that feed into the sensory (in this case, visual) areas of the brain. These top-down connections are proposed to manipulate the firing rate, or synaptic connections, of the neurons in visual areas of the brain leading to the experience of visual imagery. We can think of these top-down manipulations as creating ‘maps’ of the images that we imagine in the visual areas of our brain. When we imagine multiple images, we create multiple maps, and these maps compete for space in the brain’s visual areas. This competition and interaction between imagery maps is, perhaps, what gives rise to our capacity limits.

Why do these limitations matter? Visual imagery is, after all, involved in more tasks than buying sofas and tables from Ikea. Take the treatment of mental illnesses for example. Phobias are commonly treated by imaginal exposure. The therapy works by repeatedly exposing someone to the thing that causes them anxiety, such as spiders, plane rides, public speaking, heights etc, and this repeated exposure leads to a lessening of the fear response. For obvious practical reasons, it can be difficult to expose people to these situations repeatedly, so clinicians use imaginal exposure instead. The patient imagines the fear-inducing stimuli in as much detail as possible, and this is thought to work in a similar way to actual exposure to the stimuli.

Another form of treatment in clinical psychology involving visual imagery is imaginal rescripting, which is used to treat many different disorders such as depression, generalized anxiety, obsessive-compulsive disorder, and eating disorders. Imaginal rescripting involves participants imagining, or simulating, scenarios from the past or future that are fear- or anxiety-producing. They imagine these scenarios in as much detail as possible, and then are asked to imagine an alternative scenario or outcome that is more positive – they ‘rescript’ the memory or thought. They are also taught how to change their thinking towards these scenarios during the process.

Although imagery-based therapies, such as imaginal exposure and rescripting, have been shown to be one of the best cognitive behavioural therapies out there, they are not 100 percent effective. It might be that one of the factors influencing how well these treatments work might be the ‘realness’ of the scenarios created in the mind, which is affected by both the capacity limits of imagery, as well as the individual differences in people’s ability to create these scenarios.

In addition to therapy, we use visual imagery when we remember the past and plan for the future; when we retain and manipulate visual information in our working memories; it even plays a role in our moral judgments, and our intentions to help others. The capacity limits to visual imagery we have discovered will likely influence the amount and quality of information we are able to maintain and process during any of these situations. These limitations to our imagery can constrain what we are able to achieve, both in daily life and in therapeutic interventions.

For now, the jury is still out as to whether or not our capacity for visual imagery can be improved (indeed, that’s the question I’m working on now). But what we do know is that by studying and creating new objective ways to quantify the capacity limits of our visual imagery we can begin to understand the full limitations to the human imagination and mind, and develop new ways to overcome these.Aeon counter – do not remove

Rebecca Keogh is a postdoctoral fellow in cognitive neuroscience at the University of New South Wales in Australia.

This article was originally published at Aeon and has been republished under Creative Commons.

See What Others Are Saying

6 comments / Join the Discussion

    Rebecca, Have you considered using different imagery other than Gabor patches? Gabor patches are abstract visual symbols humans have no innate reason to remember. they have little relationship to each other. I can imagine the entire alphabet, that is much more than 4 symbols. I can images imagine arrows pointing in 8 directions. and then put them into a new image where they connect at their vertices. People can visualize many numbers of things they consider relevant. I can visualize upper and lower case letters on plaques above a school room, displayed in a school font, with the dotted lines going through the middle to show how upper and lower case characters should be written.

    some of these experiments seem to test salience and not capacity. the disconnected quality of the images likely makes them hard to visualize. Also, the lack of practical experience. an interior designer should have no problem imaging different couches and lamps in a room. consider what a fashion designer has to imagine in their head? the fabric, it’s cut, it’s weave, how it hangs, how it reflects light, how it will fit on various bodies, the color, the stitching, etc.

    you wouldn’t attempt to understand mathematical visualization or ability to using some meaningless symbols in different arrangements as a stand in for the ability to conceptualize equations. most people would only remember 3 or 4 useless symbols. but anyone doing even rudimentary algebra can remember 7 or 8 symbols and manipulate them to obtain results. (x * 3) + 2(x – 1) = 24 it contains 11 different symbols. would a person just learning algebra have a problem visualizing it and working it in their head? sure. would a student who has been doing these problems all school year? probably not. familiarity matters. an absence of familiarity does not test capacity.

    I would suggest doing visual tests with animals as the visual objects. people are very good at visualizing animals and where animals are. or food. people can visualize a table full of food from a glance with apparent ease. everyone can remember where the chocolate fountain is at a wedding, even if they are in the other room. Salience plays a large role in what is visually remembered. Gabor patches may not test capacity at all, merely capacity to visualize non-salient imagery.

    you may want to construct tests that do responses with imagery vs responses with verbal responses. I’ve noticed that many people can know where something is, indicating visual recall, but cannot explain or describe where it is. for instance in path finding. different people use different strategies. if you travelled to a new location, you may remember landmarks or you may visualize places on a map. If you recall landmarks, you likely construct a narrative to follow to arrive at a destination. If you use a map you could probably describe a map structure and a path in that structure or simply a location on the map. those are very different visual capacities. most people do some of both. but narrating what is visually recalled is different than drawing a map. You may want to use different methods of expression to suss out visual capacity. a person who maps or draws the imagery may remember differently than someone making a verbal account of what they can imagine.

    And can I suggest you hang out with sculptures, painters, and photographers (especially ones that compose or wait for specific shots). Or people who make movies. A visual filmmaker makes imagery that occurs over time. they know where things will be in an image at a point in time, and then will try to shoot it that way. they know, visually, if the take captures the visual phenomena they are trying to achieve. a sculpture sees their work from many angles, it’s weighting and structure for instance. a painter has a grasp of the motion they must make with their hand to produce an effect, such as the curve of petal ( a visual+ kinesthetic skill)

    as someone who plays with removing binocular rivalry from time to time, I’m rather convinced this is a skill that can be acquired. but it carries with it a subtle danger of seeing combinatorial objects vs a field of objects. I suspect the rivalry effect arises from the ability to see background objects through foreground obstacles. like seeing an animal though trees or brush. we select for the animal and not the obscuring branches, because the animal is the salient object and the obscuring branch is not. (conversely, if we are fleeing an animal we select for branches and not the background to choose the least obstructed path.)

    I fiddle with rivarly effects by looking at cars across a street seen through papyrus in front of my grilled bedroom window. the cars and stalks tend to overlap as I relax binocular rivalry. The same car will appear in different gaps between stalks. If i look at stalks in the foreground a single background car will appear in different gaps because the foreground image is meshing both eyes. the car appears as 2 cars because of the distance. if i look at a background car, I have binocular rivalry to see the car in a specific eye. If I relax enough to ease the binocular rivalry, the background car is simply too large to be a single car, and I interpret it as 2 cars.

    the danger with easing binocular rivalry is that it diminishes the capacity to correctly perceive background objects through foreground obstacles. it mixes up foreground and background objects as if they exist on the same plane. in practical situations you cannot separate focal distance from binocular rivalry. I suggest trying these experiments yourself. looking at back ground objects through irregular near “filters” like plants or grills or fingers.

    It’s exciting you are doing this work. very few in academia approach visual work, because so much of academic work is language based. that language bias obscures the deep visual qualities and capacities human beings have. I hope you continue to pursue this, because it’s a largely undiscovered country ripe with possible discoveries.


    It would be very interesting to see the same experiment repeated with different images instead of Gabor patches. To find out how complexity of an image and participant’s familiarity with it affects results.


    Fascinating! I have aphantasia and cannot visualize mentally at all. If you ask me to visualize a wine colored sofa I just see darkness in my minds eye. This effects my memory just like the end of your article mentions. I have a very poor memory of my past. I also have a very hard time imagining the future.


    ahh Amanda. living in the present is supposed to be the best way to engage life…Interesting to think of trauma and how it might be
    re-experienced or not…so much to think about; thank you for the contribution.

Join the discussion

Your email address will not be published. Required fields are marked *