- Recognizing specific emotions in social settings allows us to predict which behaviors are coming next, and to adapt to them.
- Researchers used video clips of emotion-laden behavior occurring in human children, dogs and monkeys.
- While human observers were generally able to identify emotional behaviors they were particularly insensitive in anticipating canine aggression.
Imagine that you are taking your dog out for its daily walk and the two of you encounter another dog. The two dogs begin to interact. Can you read the emotional context well enough to know what the dogs are about to do next? Will they start to play? Will they just go on about their own business? Most importantly, for the safety of all concerned, is one of the dogs about to attack the other?
According to a recent study from the laboratory of Juliane Bräuer at the Max Planck Institute for Geoanthropology in Germany, you may not do a very good job of anticipating a possible aggressive interaction between these dogs.
The evolutionary importance of emotions is that when we correctly interpret them during social interactions they allow us to anticipate what will happen next. This in turn gives us a chance to react appropriately. This is particularly important when the interactions are negative and one of the individuals is about to become aggressive.
A number of studies have shown that people are generally accurate in interpreting the emotions of adults from photographs. However in real life we don't have time to sit and study a fixed snapshot of an emotional expression, rather we have a few seconds to interpret dynamic expressions of emotions and predict behavior.
Watching Children, Dogs and Monkeys
This study was done by a team of researchers headed by Theresa Epperlein of the Max Planck Institute and it used brief video clips of behaviors of humans, dogs, and monkeys, socially interacting in naturalistic settings.
Data was collected from 92 participants who got to view 2 to 5 second video clips of dogs freely interacting in gardens or dog parks, or clips of children (1-10 years old) interacting in various environments, or macaque monkeys housed at Kintzheim in France during typical social interactions.
Each brief video included social cues, like facial expressions and body postures (the lead up to actual social interactions) but was stopped short before any commonly identifiable aggressive, playful, or neutral set of behaviors revealed themselves. In the case of aggression, the video clip would stop just a few frames before a dog might display a stiff body posture threat, or monkeys might give an open mouth threat, or a child might rush aggressively at the other child. In other words, this study examined how well people can anticipate what is about to happen after viewing the preliminary emotion-laden behaviors.
Half of the participants were asked to categorize the emotional tone of the interaction as aggressive, neutral, or playful. The other half were asked to predict the outcome by choosing from sentences describing three possible options, for example: one individual will attack the other, both individuals interact playfully, or they continue what they have done before without further interaction.
How Accurate Are the Judgments?
The overall results showed that participants can generally identify the emotional nature of social interactions. Specifically, although their performance was far from perfect, these observers showed an ability to describe the emotional state and foresee the subsequent behaviors for human children, dogs and monkeys. Their accuracy was statistically above chance. However, as the saying goes "the devil is in the details" and how accurately the participants predicted and categorized the situations depended on both the species and the emotional context.
The most striking result in this study was how poorly the participants did when it came to recognizing aggression in dogs. This might have been predictable since another recent study headed by Catia Correia-Caeiro of Kyoto University, used eye-tracking measures of human participants to determine what the participants in the experiment were looking at while they watched video clips of naturalistic emotional behaviors in dogs or humans.
That study found that, despite the fact that it is easier to read emotional states of dogs from their body language, the human observers focused mainly on dogs' heads and faces in the video. That means that, especially for aggressive interactions, humans were focusing on the less reliable facial signals to interpret canine emotions. The truth is that if you look in the wrong place you are not going to find the information that you need to identify emotions and predict behaviors.
Are Humans Blind to Canine Aggression?
However, in this new study there was a surprise. When it came to interpreting the likelihood of aggression, humans actually identified the reality of the situation significantly less than would be expected if they were simply blindly guessing. In clinical settings when humans recognize situations statistically less than chance, it usually is interpreted as "denial." In other words it is an indication that the person is consciously or subconsciously refusing to see and correctly identify what is going on. Are people denying the expression of aggression and the potential for hostile behavior in dogs, despite the fact that the evidence is right in front of them to view?
The authors of this study seem to have reached that conclusion since they acknowledge that "It is possible that humans are biased to assume good intentions from other humans and from 'man’s best friend', sometimes preventing us from recognizing aggressive situations in these species."
Copyright Stanley Coren, 2022.
Facebook image: Jan Mlkvy/Shutterstock
Epperlein T, Kovacs G, Oña LS, Amici F, Brauer J (2022) Context and prediction matter for the interpretation of social interactions across species. PLoS ONE 17(12): e0277783. https://doi.org/10.1371/journal.pone.0277783
Correia-Caeiro C, Guo K, Mills DS. Perception of dynamic facial expressions of emotion between dogs and humans. Animal Cognition. 2020; 23(3):465–76. https://doi.org/10.1007/s10071-020-01348-5