The emotion of shapes

Take a look at these two shapes. Which appears more "joyful"? Which appears fearful?


How about these shapes? Which is angrier? Which appears to be suffering more?


If you're like most people, the shapes that appear to be less stable (number 2 in the figures above) are also more fearful. Those that are rotated more from the vertical position (again, number 2 in the figures) are more suffering and less angry.

Assigning emotions to shapes is nothing new. In experiments as early as the 1940s, individuals have been found to consistently apply the same emotions to shapes in schematic cartoons: "angry" triangles and "loving" circles. But only one study had attempted to see if people consistently assigned emotions to static shapes based on the appearance of dynamic forces.

In that study, participants viewing two slashes ( / / ) tended to see the slash leaning towards the other as "bossy," while the one leaning away was "submissive." In 2005, Irena Pavlova, Arseny Sokolov, and Alexander Sokolov made the first study of emotions and dynamics in single shapes: they showed participants triangles, ovals, and lines in a variety of orientations, and asked them to rate the figures for the emotions they conveyed.

The triangle balanced on its point (Number 1 in the above figure) or lying on its side was viewed as the most stable, and similarly, the oval balanced on either its end or its side was seen as stable as well. Instability correlated significantly in these figures with perceived suffering and fear.

For the triangle, the oval, and the line, there was a significant negative correlation between joy and the degree of rotation from vertical: the more vertical, the more joyous the figure appeared. This correlation held for anger as well, but only in the case of the oval and the line.

Pavlova and her colleagues argue that the implied imbalance in the pictures of static objects is what leads individuals to attribute emotion to them. They note as well that neural imaging of patients with Asperger Syndrome shows that the parieto-frontal circuits in these individuals are activated differently compared to normal individuals. Since one symptom of Asperger Syndrome is an inability to detect emotional states of others, the team suggests that there is a direct link between perception of the physical orientation of an object and perception of emotional states.

This finding has implications for the field of art as well. If certain shapes are indeed associated with emotions, then this may partially explain the appeal of abstract art. Rather than being a seemingly random collection of shapes, abstract art may evoke common emotions in many viewers.

Pavlova, M., Sokolov, A.A., & Sokolov, A. (2005). Perceived dynamics of static images enables emotional attribution. Perception, 34, 1107-1116.

More like this

There's something about kids and dogs. The phrase "A boy and his dog" brings up quite a range of images: from the sweetness of Norman Rockwell to what sounds like a truly bizarre movie from 1975. Despite not being a dog-person myself (okay, not being a pet-person at all), I find the results from…
A continuation of our "greatest hits" from past Cognitive Daily postings: [originally posted on July 11, 2005] There's something about kids and dogs. The phrase "A boy and his dog" brings up quite a range of images: from the sweetness of Norman Rockwell to what sounds like a truly bizarre movie…
Have you ever played around with point-light displays? If not, take a few moments to explore the amazing site I've linked. Through these simple animated displays, we can detect gender, emotion, even species. Point-light displays have been studied for decades as a way to understand how we perceive…
I'm headed to Atlanta, for the Society for Neuroscience meeting. I don't have a poster (just presented one in France!) so I'll just be a tourist. And, hoping to run into a few neuro-bloggers like Evil Monkey, Jake of Pure Pedantry, and the Neurocontrarian. Here's how the UM Neuroscience PhD…

If I remember correctly from the very beginning of Lakoff's "Moral Politics", the whole structure (of political ideology) begins with the importance of being 'upright'.

I'd be interested to see how this test goes with people with autism.

I looked at the shapes and read the questions and honestly had now real "feeling" one way or the other about either of the shapes. I felt know "emotional" reaction in that sense from the shapes. What I did respond to was the symmetry of the shapes. I felt more "comfortable" with the upright shapes because, I believe, I was responding the the vertical symmetry. This could very well be because I am mildly autistic (Asperger's Syndrome).

I wonder if basic geometry like that is part of the facial and body language that autistics are typically blind to.

Yeah, then I read the last two paragraphs and, yup! They did indeed test people with ASD and found there were differences.

Interesting to note I suppose is the fact that I have always found abstract expressionist art to be utterly meaningless, it says nothing to me. I love art that shows strong symmetry and structure however - especially where clever things are done with that symmetry (M. C. Escher for example).

I would love to be a lab rat for some of these tests. It would be very interesting.

I find the images on the right to be more satisfying: more dynamic, adventurous, joyful than the ones on the left -- stuffy, boring.

And I do like much of modern art, at least those who play abstractly and don't aim to shock or other such nonsense. E.g., Calder.

I don't understand the researchers' last paragraph. As far a science, I do. It might be news there, and of course, it's a scientific, not anecdotal, report.

But the emotional content of the elements of art is something you pick up early in art instruction. Shape, line, color, etc. Even format. These are explicated in detail. The predominance of verticle lines versus horizontal lines versus diagonal lines in a composition can create very different emotional reactions to otherwise the same subject.

Some artists have quite elaborate, albeit idiosyncratic, element-to-emotion mapping systems, like Kandinsky.

But there are two implications in that last paragraph that are simply false. First, these emotional responses occur in representational art just as much as in abstract art. Second, abstract art is "seemingly random collection of shapes"? Some is actually, but most of it isn't. Perhaps to a novice or intuitive viewer it might seem that way. But to anyone who's paying attention, and especially to the artists, "random" is ignorant vocabularly.

Perhaps they should have wandered over to the nearest art department before offering ideas about art. There's actually knowledge therein. "...abstract art may evoke common emotions..." is barely better than "My 6-year-old kid could do that." Duh?

Come on!

By SkookumPlanet (not verified) on 27 Mar 2006 #permalink

What? I guess I don't see it. I just saw the same shape twice.

So what? It's pointing a different way.

"'random' is ignorant vocabularly"

Perhaps "random" isn't the ideal term to use there, but I did qualify it with "seemingly." My point was that abstract art has appeal even to people who don't understand the artistic conventions behind it. And the artistic conventions themselves may be based on emotional reactions that are built into the perceptual system.

Some abstract art really can be justly described as 'seemingly random' - Pollock's paintings, for example. Of course, Pollock's spashes are far from random (he was famous for the amount of thought he put into his compositions) and his works are highly emotive, but there is no use of recognizeable pattern.

Still, I concur that to anyone with any artistic background this comes across as something akin to a finding that discs make better wheels than cubes...


Sorry, I misread it. As it was over the article reference I somehow took it as part of the study they did. Today, I'm not sure how that happened.

So I read it that those authors had published that in a sci journal. I don't have as much problem with it as a blog entry. I certainly don't expect as a blogger you should go to the art department first.

I understood how you were using random. But what I'd really like to point out is that these elements work just the same in representational art. Few people realize that the same types of design considerations that become so obvious if it's abstract art, are used in representational also. Buildings become trapezoids, trees ovals and triangles, etc.

My misreading actually embarrasses me a bit.

By SkookumPlanet (not verified) on 28 Mar 2006 #permalink

No problem, Skookum.

You make a great point about representational art, and actually the authors of the study do mention its application to representational art; I was just trying to extend that application to abstract art as well.

Just as an editorial note, I should point out that we very rarely quote directly from a study, which is considered bad form in APA style. Typically, if we use a direct quote at all on Cognitive Daily, it would be to give an example of a verbal stimulus or a participant's response.

"If you're like most people...

Well ... I guess I'm not. Being asked to attribute emotional states to simple shapes struck me as pointless.

This is the sort of thing that makes me want to ask, "What answer would you like??"

By Scott Belyea (not verified) on 28 Mar 2006 #permalink

Scott, and the others who are uncomfortable with this study:

In the original study, people were asked to rate the objects along a scale for each of several possible emotions. I wonder if forced to respond in this way, if you would have made distinctions between the objects, or if you would have simply responded with a neutral level of emotion each time.

What's interesting and relevant about this study is that different individuals consistently rate the same objects with the same emotional valence, suggesting that there is some relationship between these emotional states and simple shapes.

It's certainly possible that many individuals would have no emotional associations with any shapes, but that across the population, there are still significant correlations between particular shapes and emotions.

I also find JKVisFX's reactions fascinating -- as someone who has Asperger Syndrome, he saw no emotions in the images, and also doesn't see any appeal in abstract art.

I find such physiognomic perception fascinating but agree with some of the reactions here; namely, is this forced imagination or natural perception? I find that if I suggest and subtly influence viewers they have no trouble creating and developing stories where such shapes take on meaning. It's unclear to me whether that tells us more about human imagination and not universal perceptual attribution. What emotions people see may be conditioned by the 'demand character' of the research and the cultural framework. In this regard, I wonder if much has been done on such things cross-culturally, as well as developmentally? We see movements and shapes--such as owls being wise, and the fox being clever, and sharp, jutting edges as tough, strong? What part of such attributions are independent of cultural learning? Interesting follow up to classic Heider and Simmel work.

By Gary Peterson (not verified) on 28 Mar 2006 #permalink