Fear

Do Dogs React to the Scent of Human Fear?

Can dogs be trained to be lie detectors based on their ability to scent fear?

Posted Jan 27, 2021

JackieLou DL/Pixabay
Source: JackieLou DL/Pixabay

Can dogs identify our emotional state by scent alone? Back in the 1970s I had a conversation with a colleague from the University of Utah who studied lie detection. He told me "I have been talking to some dog handlers who work for the police and they suggested to me that dogs might turn out to be good at lie detection. After all, the polygraph that we use to detect lies is simply measuring a cluster of physiological changes associated with stress. It's looking for changes in heart rate, muscle tension, breathing patterns, increased sweating and that sort of thing that might indicate anxiety and thus a lie. These police canine officers believe that their dogs know when someone is lying. They reason that this is because when people are fearful and stressed over the possibility that their guilty behavior might be discovered, their body chemistry might change a wee bit, enough at least to alter the odor of their sweat. Since dogs are so good at scent discrimination perhaps they could be taught to respond to these chemical signals and to point out to us that a person is likely to be telling us lies."

To the best of my knowledge there were no published reports exploring this question coming out of his laboratory; however some recent research from a team headed by Biagio D’Aniello of the Department of Biology at the University of Naples Federico II suggests that this area of inquiry might be worth a second look. This is because one of the necessary components for canine lie detection, namely the ability of dogs to detect fear or stress by scent alone, now appears to have been confirmed.

Obviously, if you are going to explore the behavioral responses of dogs to emotional scents, the first thing you need are appropriate samples of emotion-laced odors to test. To obtain these the investigators had eight different Caucasian male university students watch videos which were designed to evoke fear and stress or, alternatively, videos designed to produce feelings of happiness and relaxation. During the video viewing period the students had scent collecting pads placed under their armpits to collect their odors and sweat. These pads were later frozen and preserved to serve as scent samples. Prior to testing each pad was cut into quarters and one quarter was put in a container along with samples from three other individuals who had experienced the same emotional condition. This gives you a sort of "super sample", which in effect averages the fear or the happiness scent of four people and smudges out the other differences unique to particular individuals. In addition there was a control scent, obtained by using pads fresh out of the pack which had not been exposed to any biological sources of odor.

The testing set-up was fairly straightforward: 84 dogs and their owners were individually tested in a room which contained two chairs diagonally opposite each other in the corners. In one of these the dog's owner sat, while in the other was a stranger. Both remained passively seated during the testing session. In the center of the room was an apparatus which contained the odor sample. After the odor sample (the fear scent, happy scent or control scent) was released into the air the dog's behavior was videotaped and analyzed.

The researchers were looking for a number of different things. Obviously they were looking for signs of stress, such as panting, yawning, mouth licking, barking, pacing, and so forth. They also noted whether the dog was looking toward or approaching the door which allowed exit from the room. In addition they monitored which person the dog approached and paid attention to — its owner or the stranger — and which person the dog would try to interact with. Previous research has shown that dogs tend to hover around their owner when they are feeling stressed, using that familiar person as a safe base.

The major finding was quite unambiguous: When the odor released into in the room was based on scents collected from people experiencing fear or anxiety producing conditions, the dogs indicated that they recognized something negative by showing more signs of stress and staying close to and trying to interact more with their owners. Both the male and female dogs responded in the same way to the fear-based odorant.

There was a surprise, however, and this had to do with responses to the scent associated with happiness. Researchers found that the male dogs really didn't seem to recognize, or at least didn't respond to, the happiness scent in any way that was different than how they responded to the emotion-free control scent. Female dogs, however, did seem to recognize the happiness scent, since when it was in the air they showed an increased interest in the stranger on the other side of the room. Thus both male and female dogs responded to the scent of fear, while only the females seemed to be responsive to the scent of happiness.

The fact that dogs can recognize the scent of fear or stress, and respond to it, does seem to reopen the question as to whether dogs can be used to detect whether a person is lying or telling the truth.

While this research suggests that all dogs might have some ability to become biological lie detectors, this data also suggests that female dogs also have the ability to detect and respond to scents associated with happy emotions. Thus it might be that female dogs can be assigned an additional, non-forensic task, namely detecting whether or not a person is experiencing positive emotions in the presence of their owner — perhaps a nice addition to a dating app?

Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission.

Facebook image: moj0j0/Shutterstock

References

D’Aniello, Biagio, Fierro, Barbara, Scandurra, Anna, Pinelli, Claudia, Aria, Massimo and Semin, Gün R. (2021). Sex differences in the behavioral responses of dogs exposed to human chemosignals of fear and happiness. Animal Cognition, https://doi.org/10.1007/s10071-021-01473-9