When a neutral face isn't neutral

i-eca0cf2af9fc3ac4445c7dff7d8aab70-research.gifThe Kuleshov Effect, discovered nearly a century ago by Soviet filmmaker Lev Kuleshov, posits that the context in which we see an image of an actor's face will determine the emotion the face portrays. For example, take a look at this short little clip I made (QuickTime required). First you'll see a gray screen, then a photo, then a second gray screen, and another photo of a face, taken just after that person looked at the first photo:

What emotion would you say characterizes the second face? Is it neutral, subtly happy, or subtly sad?

Kuleshov's work suggests that most viewers will see that second face as happier in the context of the happy photo preceding it, compared to if they had seen it on its own. If the identical photo had been preceded by a negative image such as an aggressive dog, people would rate it more negatively.

Research during the 1980s and 1990s confirmed the effect, but it has been exploited by filmmakers countless times during the intervening decades -- the neutral face of a heroine is seen as sad if she's just witnessed her lover's death, but happy if she's anticipating his arrival on the next train.

Most recently, a team led by Dean Mobbs replicated the Kuleshov Effect while viewers' brain activity was monitored via fMRI. Mobbs' team showed movies similar to the example above to 14 volunteers. Sometimes the first (context) image was neutral, sometimes it was positive (happy), and sometimes negative (fearful). While the fMRI monitored brain activity, viewers rated the second (neutral) image as positive or negative. This chart shows the results:


The emotion of the neutral face was rated as significantly more positive in a positive context than in a neutral or negative context. The reverse occurred for a negative context. The two brain scans above show one slice of brain activity while viewing the identical face in a positive and negative context. As you can see, there was significantly more activity in the anterior cingulate cortex (ACC) in the negative context than in the positive context. This makes sense, since the ACC has been found to play a role in emotional appraisals.

Mobbs et al. found differences in a variety of brain regions, all while the viewers were rating the identical face. The team also asked viewers to rate faces that showed subtly positive or negative expressions, and again found that context influenced both the ratings and brain activity.

Mobbs and his colleagues argue that this research shows that there is a neurobiological basis for the Kuleshov Effect.

Mobbs, D., Weiskopf, N., Lau, H.C., Featherstone, E., Dolan, R.J., & Frith, C.D. (2006). The Kuleshov Effect: The influence of contextual framing on emotional attributions. SCAN, 1, 95-106.

More like this

I don't know. I didn't see the second face as any more positive in the context of the first photo. If anything, I thought the person in the second photo looked angry.

Me too. He looks like the pissed off older brother.

A neurobiological basis? As opposed to what, dualism?

I hate fMRI studies like this. We already knew it happened in the brain.

I agree with Sadie and Rachel. The second picture appeared upset or aloof. I think perhaps the person's expression is not actually neutral enough for the example.

Actually, I think he seemed pissed-off because we expect him to look at least a little happy in response to those happy kids. The gap between our expectation and his actual expression makes him look more negative than without the photo.

By Andy James (not verified) on 22 Mar 2007 #permalink

I thought the second person looked mad, too. Could it have anything to do with the fact that the second picture is up for only a fraction of the time as the first picture?

By Katie Gill (not verified) on 22 Mar 2007 #permalink

I'm not sure about that either... Russell & Fehr (1987, J. Exp. Psych.) used Paul Ekman's original photographs and found that the second, neutral face looked happier after a sad face but sadder after a happy face. After seeing a happy face the neutral face looks sadder in contrast.

In other words, there are contrast effects rather than carryover effects.

I think all this demonstrates is that context plays a role on our judgements of other peoples emotional states. When given a emotionally neutral face we natural look for other external cues to determine their emotional state. Was the study controlled by showing neutral picture followed by faces with obvious emotional reactions?

Expectations can affect how we interpret an expression. Seeing somebody doing something they're not supposed to enjoy will lead us to seeing any expression of enjoyment as one of displeasure or dismy. Or we can see an expresson aimed at one element as being aimed at another.

In short, we tend to see what we want to see. Good old observer bias. We know that humans are the only large free-range primates found in North America, so when we see a large free-range non-human primate in a North American setting our first assumption is that it has to be a man in some sort of costume.

Then you have what I call the "Righteous Cause Effect". Holding on to a conclusion or observation in the face of all evidence against it, because you know it's the right thing to do.

Not pissed but lethargic at best. I think we the 21 century viewer might be a bit more sophisticated than early Soviet viewers a century ago.