facial feedback hypothesis simple definition

Facial Feedback Hypothesis (Definition + Examples)

We show our emotions through our facial expressions. We smile when we are happy and frown when we are angry. This is one of the ways we communicate our feelings to others. But did you know it might also work the other way around? Our facial expressions can influence our emotions. 

This is the main assumption of the facial feedback hypothesis. 

What Is the Facial Feedback Hypothesis?

The facial feedback hypothesis suggests that contractions of the facial muscles communicate our feelings not only to others but also to ourselves. In other words, our facial movements directly influence our emotional state and our mood even if the circumstances around us don't change!

All humans are thought to share seven basic emotions : happiness, surprise, contempt, disgust, sadness, anger, and fear. Each one of these emotions has unique facial expressions associated with it. Raised lip corners and crow’s feet wrinkles around eyes mean joy, while tightened lips and eyebrows pulled down signify contempt. 

But facial expressions are more than just representations of our emotions. They contribute to and sustain what we are feeling. 

Example of Facial Feedback Hypothesis at Work

The best example of this theory is easy to perform. Go to the mirror and smile. Keep smiling...keep smiling! Even if you were in a bad mood before, you are likely to lighten up and maybe even start laughing! (This is much more fun to try than scowling!)

Who First Wrote About Facial Feedback Hypothesis?

The origins of facial feedback hypothesis can be traced back to the 1870s when Charles Darwin conducted one of the first studies on how we recognize emotion in faces. Darwin suggested that facial expressions of emotions are innate and universal across cultures and societies. In his book The Expression of the Emotions in Man and Animals, he argued that all humans and animals show emotion through similar behaviors. 

Paul Ekman's Contributions to Facial Feedback Hypothesis

Numerous studies have since confirmed Darwin’s idea that facial expressions are not socially learned. Instead, they appear to be biological in nature. In the 1950s, American psychologist Paul Ekman did extensive research on facial expressions in different cultures. His findings were in line with Darwin’s idea of universality. Even the members of most remote and isolated tribes portrayed basic emotions using the same facial movements as we do.

What’s more, expressing emotions through facial movements is not any different in people who were born blind. Although they can neither see nor imitate others, they still use the same facial expressions to project their emotions as sighted people do. 

There are, however, a few exceptions. 

People with schizophrenia and individuals on the autism spectrum have not only difficulty recognizing nonverbal expressions of emotions, but also producing these spontaneous expressions themselves. They typically either remain expressionless or have looks that are hard to interpret.

The James-Lange theory of emotion

A decade after Darwin’s study, the father of American psychology William James and Danish physiologist Carl Lange proposed a new theory of emotion that has served as a basis for the facial feedback hypothesis. ​ The James-Lange Theory of Emotion implies that our facial expressions and other physiological changes create our emotions. 

James famously illustrated this assertion with a story of a man being chased by a bear. A man is unfortunate enough to encounter a bear in a forest. He is afraid and, naturally, his heart races and he is sweating as he starts running away. According to the psychologist, it is precisely these physiological changes that provoke the man’s feeling of fear. In other words, he doesn’t run from the bear because he is afraid. He is afraid because of his physiological response to running away. 

Fritz Strack’s cartoon experiment

In 1988, German psychologist Fritz Strack and his colleagues conducted a well-known experiment to demonstrate the facial feedback hypothesis. The participants in Strack’s experiment were instructed to look at cartoons and say how funny they thought these cartoons were. They were asked to do this while holding a pen in their mouths. Some participants held the pen with their lips, which pushed the face into a frown-like expression. Others held it with their teeth, forcing a smile. 

Strack’s results were in line with the facial feedback hypothesis and were since confirmed by several other studies. The participants who used a pen to mimic a smile thought that the cartoons were funnier than those who were frowning. The participants’ emotions were clearly influenced by their facial expressions. 

Characteristics of Facial Feedback

The brain is hardwired to use the facial muscles in specific ways in order to reflect emotions. When contracted, facial muscles pull on the skin allowing us to produce countless expressions ranging from frowning to smiling, raising an eyebrow, and winking. In fact, we are capable of making thousands of different facial expressions, each one lasting anywhere between ​ 0.5 seconds (microexpressions) to 4 seconds. 

But facial expressions can indicate various degrees of emotions as well. When we are slightly angry, we display only a light frown and somewhat furrowed eyebrows. If we are furious, our expression becomes more distinctive. In addition, we can show combinations of different emotions through subtle variations of our facial movements.

The facial feedback hypothesis has the strongest effect when it comes to modulation, that is, intensifying our existing feelings rather than initiating a completely new emotion. 

Modulating also means that if we avoid showing our emotions using our facial muscles we will, as a consequence, experience a weaker emotional response. We won’t feel the emotions as strongly as we otherwise would. The lack of facial expressions or inhibition of these expressions lead to the suppression of our emotional states. 

Applications of the Facial Feedback Hypothesis

The facial feedback phenomenon has several possible applications. It can help us be more positive, have better control of our emotions, and strengthen our feelings of empathy. We can simply use the facial feedback hypothesis to make us feel better in situations that we would rather avoid. If we force a smile instead of frowning at a boring event, for example, we may actually start to enjoy ourselves a bit more. We can use the same exercise whenever we are feeling overwhelmed, powerless, or stressed. 

Research shows that regulating emotions through facial feedback can have positive outcomes in areas ranging from psychotherapy to child education and endurance performances.

Related posts:

• Paul Ekman Biography - Contributions To Psychology
• Body Language Basics - How to Read Someone
• Facial Expressions of Emotions (Microexpressions)
• James-Lange Theory of Emotion (Definition + Examples)
• Two Factor Theory of Emotion

Reference this article:

About The Author

Free Personality Test

Free Memory Test

Free IQ Test

PracticalPie.com is a participant in the Amazon Associates Program. As an Amazon Associate we earn from qualifying purchases.

Follow Us On:

Youtube Facebook Instagram X/Twitter

Psychology Resources

Developmental

Personality

Relationships

Psychologists

Serial Killers

Psychology Tests

Personality Quiz

Memory Test

Depression test

Type A/B Personality Test

© PracticalPsychology. All rights reserved

Privacy Policy | Terms of Use

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List

How the Experience of Emotion is Modulated by Facial Feedback

Sven söderkvist.

Department of Psychology, Uppsala University, Box 1225, 751 42 Uppsala, Sweden

Kajsa Ohlén

Ulf dimberg.

The facial feedback hypothesis states that facial actions modulate subjective experiences of emotion. Using the voluntary facial action technique, in which the participants react with instruction induced smiles and frowns when exposed to positive and negative emotional pictures and then rate the pleasantness of these stimuli, four questions were addressed in the present study. The results in Experiment 1 demonstrated a feedback effect because participants experienced the stimuli as more pleasant during smiling as compared to when frowning. However, this effect was present only during the critical actions of smiling and frowning, with no remaining effects after 5 min or after 1 day. In Experiment 2, feedback effects were found only when the facial action (smile/frown) was incongruent with the presented emotion (positive/negative), demonstrating attenuating but not enhancing modulation. Finally, no difference in the intensity of produced feedback effect was found between smiling and frowning, and no difference in feedback effect was found between positive and negative emotions. In conclusion, facial feedback appears to occur mainly during actual facial actions, and primarily attenuate ongoing emotional states.

Introduction

The facial feedback hypothesis is based on the idea that a facial expression not only expresses an emotion, but also that expression and experience are linked in that afferent sensory feedback from the facial action influences the emotional experience. Elevating your cheeks can make you happier, just as furrowing your brow can make you angrier. The idea was introduced by Darwin ( 1872 ) when he noted that the experience of an emotion seemed to be intensified when the emotion was freely expressed, but softened when repressed. James ( 1884 ) also presented ideas along the same lines, but it took almost a century before the specific facial feedback hypothesis was formulated by Tomkins ( 1962 ), Gellhorn ( 1964 ), Izard ( 1971 ), Ekman ( 1973 ) and Buck ( 1980 ). Recently, Niedenthal ( 2007 ) and (Niedenthal et al. 2009 , 2010 ) provided the theoretical framework of embodied emotion to explain, among other things, afferent feedback from facial actions. Empirical research has found that facial feedback has the ability to both modulate present emotions and to initiate emotions (for reviews see Adelmann and Zajonc 1989 ; Buck 1980 ; McIntosh 1996 ). This has been firmly established for happiness and anger and the corresponding facial expressions (e.g., Dimberg and Söderkvist 2011 ; Laird 1974 ; Rutledge and Hupka 1985 ). Also, feedback mechanisms seem involved in the expression of fear and sadness (e.g., Flack et al. 1999 ), as well as for surprise and disgust (e.g., Lewis 2012 ).

Recent facial feedback research has sought to improve knowledge of basic mechanisms of facial feedback, and attempted to demonstrate clinical benefits from manipulating the facial muscles. One insight to the basic mechanisms has been provided by Hennenlotter et al. ( 2009 ), who demonstrated that facial feedback modulates neural activity in the amygdala. Using fMRI they found that amygdala activation was attenuated during imitation of angry facial expressions in participants with Botox-induced paralysis of their corrugator muscles, the muscles used when frowning. Kim et al. ( 2014 ) further demonstrated that facial feedback modulates activity in the amygdala in an fMRI study where participants were passively exposed to happy and angry facial expressions. Amygdala activity was attenuated when corrugator muscles were paralyzed by Botox, as compared to tests before injections and also to tests after the effects had worn off.

Furthermore, Mori and Mori ( 2010 , 2013 ) have demonstrated significant effects from a passive form of facial feedback, where facial expressions were formed, without any muscle action, on the face of participants with the help of elastic adhesive bandages. Interestingly, since no muscles were activated, any feedback must have originated from sensations in the skin or from proprioceptive signals in the cutaneous muscles. Moreover, Dimberg and Söderkvist ( 2011 ) demonstrated that facial feedback was more effective at modulating emotions, as compared to initiating emotions, and that feedback was equally effective at modulating both positive and negative emotions. That study also found that effects of facial feedback actions could be detected several minutes later, a result that calls for further research.

Clinical benefits from manipulating facial muscles have been demonstrated in several studies, where Botox injections were used to paralyze the corrugator muscles of clinically depressed patients. The results showed significant mood improvements in patients, and these effects were attributed to reduced facial feedback from the corrugator muscles (Finzi and Rosenthal 2014 ; Magid et al. 2014 ; Wollmer et al. 2012 , 2014 ).

Yet, it remains unclear how facial feedback actually works. The most commonly held theory states that when certain facial muscles are activated, afferent feedback from muscular proprioceptive patterns activates corresponding affect programs (Adelmann and Zajonc 1989 ; Tomkins 1962 ). However, sensory feedback from the facial skin has also been proposed as a possible contributor to facial feedback (Tomkins 1980 ). Furthermore, the vascular theory of emotional efference provides another explanation suggesting that facial actions can regulate brain blood temperature by altering nasal air flow, which in turn could produce changes in emotional state (Zajonc et al. 1989 ). However, it is inherently difficult to identify the neurological basis of the above theories and it also follows that it is difficult to specifically test the theories. Alternatively, one way forward could be to further explore different effects of facial feedback and in this way improve the description of how facial feedback effects occur and thereby extend the boundary conditions of the hypothesis.

Therefore, in two experiments the present study addressed a series of questions that together could provide a more detailed description of how facial feedback effects occur. The two experiments used the voluntary facial action technique, which was developed in Dimberg and Söderkvist ( 2011 ), based on the method used in Dimberg et al. ( 2002 ). In this paradigm participants are told a cover story that the purpose of the experiment is to measure the reaction times of different facial muscles. The cover story has proved effective in hiding the true purpose of the manipulation, which in fact is to get the participants to perform particular facial actions associated with specific emotions. In facial feedback experiments it is common to use a cover story in order to lead the participants’ attention away from facial expressions and emotions to ensure that they are unconsciously affected by the feedback, which also serves to avoid demand characteristics. In the present paradigm participants are then required to react with their zygomatic muscles, which are the muscles used when smiling, and/or corrugator muscles, which are used when frowning, to different kinds of pictures, which encourage facial actions that resemble naturally occurring facial expressions. When testing the modulating ability of facial feedback, emotion evoking stimuli, such as pictures of happy and angry faces, have been successfully used (Adelmann and Zajonc 1989 ; Dimberg and Söderkvist 2011 ; McIntosh 1996 ). While reacting to the pictures, the participants also rate how pleasant/unpleasant they find them, and after completion each participant has thus smiled one time and frowned one time at each picture. If they rate the pictures as more pleasant and less unpleasant when they smile, as compared to when they frown, a general facial feedback effect has been demonstrated.

In Experiment 1, a question first raised by Dimberg and Söderkvist ( 2011 ) was further investigated, and focused on whether facial feedback actions only affect us when they are performed, or if the effect also remains to affect us later.

In Experiment 2, three questions were addressed. First, does facial feedback primarily enhance or attenuate present emotions? We know that facial feedback can both enhance and attenuate emotions, but it has not previously been specifically investigated if feedback is more effective in modulation where the facial action is congruent with the present state to enhance it, or if it is more effective in incongruent modulation to attenuate the emotion. The second question was if smiling or frowning generally produces a stronger feedback effect than the other, or if they produce equally strong feedback effects. It is well established that these two major facial actions both produce facial feedback, but it remains unclear if they are equally effective. Third, another question that was also first addressed in Dimberg and Söderkvist ( 2011 ) was if positive or negative emotions are easier to modulate with facial feedback or if both are modulated equally effective by the feedback.

Experiment 1

The aim of the first experiment was to further investigate if the effect of facial feedback actions could remain and also affect us later. In Dimberg and Söderkvist ( 2011 ), a significant effect of facial feedback was found about 4 min after the facial action. The design in that experiment, based on the voluntary facial action technique, let participants in a first phase react with smiles/frowns to pictures with positive or negative valence, without rating the pictures. In a second phase they viewed the pictures once again, and this time they did not voluntary move their facial muscles but just rated how pleasant/unpleasant they found the pictures. A facial feedback effect was demonstrated, because the rating of a picture in the second phase was affected by which facial action the participant had performed to that picture in the first phase. That is, participants found pictures to be more pleasant if they had smiled to them 4 min earlier, than if they had frowned.

However, the experiment by Dimberg and Söderkvist ( 2011 ) had some limitations. Specifically, no effects were measured during the facial actions in the experiment, and no effects later than 4 min were evaluated. It is therefore unclear if the intensity of the effect had changed during the 4 min, and it is also unclear for how long the effect remained. In order to investigate this question more extensively, the present Experiment 1 was designed to measure the facial feedback effects both during the facial actions, 5 min after the actions and approximately 1 day after the actions. The procedure and cover story in the present experiment was essentially the same as the above presented voluntary facial action technique, but some changes were introduced in order to facilitate the measurement of effects at the three different points in time.

If feedback effects would be detected after the initial action, they could be explained in terms of embodied emotion (e.g., Niedenthal 2007 ) and/or evaluative conditioning (e.g., Hofmann et al. 2010 ). Considering the empirical literature on facial feedback, a feedback effect during the facial actions was expected. Based on the results in Dimberg and Söderkvist ( 2011 ), a feedback effect after 5 min could be expected as well. However, we had no specific expectations for the outcome after 1 day since this has not been investigated before.

Participants

Thirty-two persons (mean age = 25.40, SD  = 5.20), balanced by gender, participated in the experiment and most of them were students at Uppsala University. No one that studied psychology or previously had participated in similar studies was included. Five persons were excluded from the study due to not following the instructions correctly, and they were replaced by other participants. As compensation, participants were given three movie vouchers at a total value of 300 SEK (approximately 30 USD).

Experimental Design

Because the aim of the experiment was to examine facial feedback effects over time, the two main independent variables were muscle (smile vs. frown) and time (during action vs. after 5 min vs. after 1 day). A secondary independent variable was stimulus (happy vs. angry). Thus, the experiment had a within subject 2 × 3 × 2 factorial design. The dependent variable was the ratings of how pleasant/unpleasant participants experienced stimuli.

Much of the apparatus served only a deceptive purpose in order to strengthen the cover story. The psychophysiological laboratory, for example, played an important role for the credibility of the cover story. The participants were individually tested in a small room inside the laboratory, where they were seated in a comfortable chair with a 17 inch computer monitor placed 70 cm in front of them. A PC placed outside the room was used to display the pictures. In accordance with the cover story, electrodes to ostensibly measure EMG were attached over the corrugator supercilii (frowning) and the zygomatic major (smiling) muscle regions.

The stimulus material consisted of six pictures of happy faces and six pictures of angry faces, selected from Ekman and Friesen’s Pictures of facial affect ( 1976 ). Even if some studies (e.g., Wangelin et al. 2012 ) have demonstrated that pictures of facial expressions do not typically induce strong emotional engagement, other studies performed with similar exposure techniques and experimental situations (e.g., Dimberg 1982 , 1990 ; Dimberg et al. 1998 , 2000 , 2002 ) have demonstrated that pictures of facial stimuli can reliably induce positive and negative emotions of equal intensity. One further reason for using these stimuli was to maintain consistency with Dimberg and Söderkvist ( 2011 ), where the same material were used. The pictures in the present study were presented in the form of PowerPoint slideshows, where each picture was shown for 8 s and then followed by a black screen for 20, 25 or 30 s before the next picture. Participants rated picture experience using the two categories pleasantness and unpleasantness on a scale from 0 to 100 ( not at all to very much ). The reason for using two opposing scales was that we could not rule out the possibility that they would be differently sensitive to the different muscle and stimulus conditions in the experiment. For each picture, a sheet containing both scales was used.

As part of a distraction task, performed between experimental phases in order to prevent carry-over effects, participants pressed a button on a handheld device to seemingly register their reaction time in response to displayed pictures of circles and triangles, three of each.

Cover Story

The cover story was essentially the same as in Dimberg and Söderkvist ( 2011 ), with a couple of small adjustments for the new procedure. Information on signup posters and in scheduling emails had briefed the participants that the experiment would measure reaction times in different muscles and to different kinds of stimuli. At arrival, they were given a more detailed introduction to the experiment, including a presentation of the psychophysiological apparatus. The experiment was presented as the latest in a run of experiments that aimed to clarify how fast human beings could react. The claimed purposes of this specific experiment were to (first) compare the reaction times in facial muscles with the more traditional reaction time measurement of pressing a button with your thumb, and (second) to see how much the reaction time varied over different days. The stimulus material was presented as pictures of human faces and of geometrical figures and the reason for using them was to compare with previous experiments that had used other kinds of stimuli, like audio signals. Participants were told that the reason for rating the pleasantness and unpleasantness of the stimuli was to control for the possibility that the reaction time is affected by how we like the picture. The purpose of this explanation was to mislead participants that might suspect a hidden agenda into believing that the agenda was in fact to see if different kinds of pictures affected the reaction time.

Participants attended the laboratory on three separate days within the same workweek. They performed two nearly identical cycles of testing (congruent/incongruent reactions) and in each cycle they rated the stimuli at three different points in time (during action, after 5 min, after 1 day). Half of the participants began with the congruent reactions, and the other half began with the incongruent reactions. In the congruent cycle they were instructed to react by elevating their cheeks as fast as possible if the displayed picture was of a happy person, and to contract their eyebrows as fast as possible if the picture was of an angry person. In the incongruent cycle they received the reversed instructions and thus elevated their cheeks to angry persons and contracted their eyebrows to happy persons. Consequently, and as can be seen in Fig.  1 , the procedure consisted of six different phases: 1—congruent (or incongruent) reactions, ratings during action, 2—no reactions, ratings after 5 min, 3—no reactions, ratings after 1 day, and then a new cycle started with 4—incongruent (or congruent) reactions, ratings during action, 5—no reactions, ratings after 5 min, and 6—no reactions, ratings after 1 day. Additionally, phases performed during the same day had distraction tasks between them to prevent carry-over effects.

The procedure over three separate testing days for each participant

After the initial presentation of the experiment, the first phase began (congruent or incongruent reactions, ratings during action). The participants were asked to sit down in the smaller room where the electrodes were attached on their face. Half of them were randomized to receive the congruent instructions for how to react, while the other half received incongruent instructions. All were told that in order for the apparatus to register their reaction time properly they needed to make a distinct muscle contraction and keep it contracted for a couple of seconds. Finally, they were introduced to the rating scales and instructed to rate for each picture, during the display, how pleasant and unpleasant they experienced it. They were then left alone and exposed to all the six happy and six angry pictures.

After this first phase, a 4 min distraction task followed where they first received instructions for how and when to use the handheld device. Half of the participants were instructed to press the button if the displayed picture was of a circle, and to do nothing if it was of a triangle. The other half received the opposite instructions, to press the button if the picture was of a triangle and to do nothing if it was of a circle. During this task they did not rate the displayed pictures. All subsequent distraction tasks had similar structure.

After the distraction task, the second phase (no reactions, ratings after 5 min) followed. The participants received instructions that they would see the pictures of human faces once again, but this time their reaction time would not be measured. Instead they were instructed to keep their facial muscles still and relaxed. New rating sheets were handed to them together with a repetition of the rating instructions. The stated reason for this extra rating session was to ensure stable and valid measurements. Participants were then exposed to the pictures in a different order as compared to the first phase.

When the participants returned for day two, the third phase began (no reactions, ratings after 1 day). They were again asked to sit down in the smaller room and the electrodes were attached. They were then informed that the day would begin with a rating session similar to the one that ended day one, where they did not react with their facial muscles. Rating sheets were handed out and they were exposed to the pictures in a different order compared to the first and second phase.

After a distraction task, the fourth phase began (incongruent or congruent reactions, ratings during action) and it was identical to the first phase, but with the critical muscle instructions reversed. After another distraction task, the fifth phase followed (no reactions, ratings after 5 min), which was identical to the second phase.

Just as in previous days, all participants were asked to sit down in the smaller room while electrodes were attached. This sixth and final phase (no reactions, ratings after 1 day) was identical to the third phase, and picture ratings were performed with muscles relaxed.

After the rating session, participants were informed that they had completed the experiment. They were then interviewed in order to determine if they had followed the instructions correctly, and if they at any point had realized the true purpose of the experiment, which none of them had. Finally, they were informed about the true purpose and were asked to keep it secret in conversation with others.

Transformations and Analysis

As a result of the experimental procedure, each participant had rated how pleasant and unpleasant they found the six happy and six angry pictures during smiling respectively frowning within each of the three different conditions of time (during action, after 5 min, after 1 day). Similar to the procedure in earlier studies (Dimberg and Söderkvist 2011 ) we compared the pleasantness and unpleasantness ratings in a preliminary analysis to determine if there were any major differences between them. When we correlated the two ratings within each participant the median correlation coefficient for all participants was −0.85 ( SIQR  = 0.11), indicating that the two measures seemed to reflect the same emotional quality. Furthermore, the results provided by the pleasantness and the unpleasantness ratings pointed in the same direction and the interpretations of them were quite similar. Consequently, to provide a more distinct presentation of the results we therefore merged the two ratings into a single bipolar rating score by subtracting the unpleasantness ratings from the pleasantness ratings. The resulting bipolar pleasantness rating scale ranged from −100 to 100, with a positive value thus indicating pleasantness, and a negative unpleasantness.

Because the purpose of the study was to measure and compare the magnitude of facial feedback effect at different points in time a new dependent variable specifically representing this was created. The facial feedback effect ( FFE ) score was calculated for each rated picture by subtracting the frown rating from the corresponding smile rating within each condition of time. A positive value in the resulting score thus indicated a facial feedback effect, since the picture had been rated as more pleasant during or after the smile, as compared to during or after the frown. Also, for each participant, the ratings of the six happy and the six angry pictures within each condition of time were collapsed into one mean for the happy and one mean for the angry pictures. Consequently, each participant had six mean FFE scores, with one score during positive emotions (to happy pictures) and one score during negative emotions (to angry pictures) for each of the three conditions in time. To analyze the data a 3 × 2 within subject ANOVA was performed, with the factors time (during action vs. after 5 min vs. after 1 day) and emotion (positive vs. negative). The significance level was set to 0.05 for all tests.

A first analysis revealed that there was no main effect of emotion and no interaction between emotion and the time factor, both F :s < 1. Therefore, to simplify the analysis and the presentation of results, the data for positive and negative emotions were collapsed so that in the final analysis only the mean FFE scores as a function of time are reported.

There was a significant main effect of time, F (2,62) = 4.57, MS e  = 147.12, p  < 0.05, η p 2  = 0.13, and as can be seen in Fig.  2 this was due to a much higher mean FFE score during action, as compared to the scores after 5 min and after 1 day. Subsequent t -tests further confirmed that the mean FFE score during action [ M  = 7.42, SD  = 19.28, 95% CI (0.47, 14.37)] was significantly higher than the score after 5 min [ M  = −1.01, SD  = 12.84, 95% CI (−5.64, 3.62)], t (31) = 2.78, p  < 0.05, d  = 0.69, and after 1 day [ M  = 0.08, SD  = 9.40, 95% CI (−3.31, 3.47)], t (31) = 2.42, p  < 0.05, d  = 0.61. Additional one-sample t -tests also demonstrated that only the mean FFE score during action differed significantly from zero, t (31) = 3.46, p  < 0.05, d  = 0.61, while the mean FFE scores after 5 min and after 1 day did not differ from zero, both t :s < 1. That is, facial feedback effects were demonstrated only during action, but no remaining effects during the two later phases.

Mean facial feedback effect (FFE) score during action, after 5 min, and after 1 day, with error bars representing one standard error

The aim of the experiment was to investigate if the effect of facial feedback actions remained to affect us at a later time. As expected, and consistent with earlier findings (Adelmann and Zajonc 1989 ; Dimberg and Söderkvist 2011 ; McIntosh 1996 ), there was a significant facial feedback effect, with a medium to large effect size, during the facial action. However, there were no effects after 5 min or 1 day. Thus, the present experiment did not replicate the result of the previous experiment by Dimberg and Söderkvist ( 2011 ), where an effect of facial feedback was detected about 4 min after the facial action. The two experiments had slightly different designs though, where the present experiment measured the effect at three different points in time and also included specific distraction tasks. It is possible that these differences could account for the diverging results, and the most likely explanation would then be that any remaining effects were abolished by the distraction tasks. If that is the case, it would suggest an initially strong and influential effect of facial feedback, but an effect that is also short-lived and easily dissipated once the facial action is over. The present results interpreted in terms of embodied emotion (e.g., Niedenthal 2007 ) would suggest that the reactivation of the facial action performed at the first exposure was not strong enough to affect the experience at the later exposures. Similarly, the results interpreted in terms of evaluative conditioning (e.g., Hofmann et al. 2010 ) would suggest that the affective signature tied to the stimulus by the facial action was not strong enough to affect the evaluation of the stimulus at later exposures.

As compared to the original voluntary facial action technique, the procedure in the present experiment spanned over several days, which presented a challenge for the cover story. Whereas participants in previous experiments only had to believe the cover story for an hour, participants in the present experiment had to believe it for half a week, and this of course increased the risk that they would realize the true purpose. However, not a single participant realized the true purpose, which further confirms that the cover story used in the voluntary facial action technique is robust and can be used in different experimental approaches and designs.

In conclusion, the results in the present Experiment 1 suggest that the effect of facial feedback primarily occurs at the same time as the facial action is performed.

Experiment 2

In Experiment 1, there was a significant effect of facial feedback during the facial actions. However, similar to many previous studies (Adelmann and Zajonc 1989 ; Dimberg and Söderkvist 2011 ; McIntosh 1996 ), this feedback effect was calculated as the difference in ratings between smiling actions and frowning actions. That is, without any ratings during a neutral action, to which the smiling and frowning ratings could be separately compared, it is not possible to conclude to what degree this feedback effect was the result of enhancing or attenuating modulation, and if smiling or frowning produced stronger effect than the other. Therefore, Experiment 2 introduced a neutral action (see below) with the aim to address these questions and provide a more detailed description of how facial feedback effects occur. The experiment addressed three different questions and they will first of all be presented in more detail.

The first question was if facial feedback primarily enhances or attenuates a present emotion. For example, with a positive emotion present, such as during exposure to pictures of happy faces, smiling could further enhance positivity, while frowning could attenuate it. And in contrast, with a negative emotion present, such as during exposure to pictures of angry faces, smiling could attenuate negativity, while frowning could further enhance it. That is, enhancing modulation consists of both smiling during positive emotions and frowning during negative emotions, while attenuating modulation consists both of frowning during positive emotions and smiling during negative emotions. In order to determine if facial feedback primarily enhances or attenuates emotions the magnitude of facial feedback effect in the enhancing respectively the attenuating modulation need to be measured and compared, which means that we need to determine the specific feedback effects of smiling and frowning. As mentioned above, with the help of a neutral muscle condition these effects can be calculated by separately comparing the ratings when smiling and the ratings when frowning to the ratings in a neutral condition. For example, there would be a specific feedback effect of smiling if the ratings when smiling were higher than the ratings in a neutral condition, since feedback from a smile is expected to increase the ratings of pleasantness. On the other hand, there would be a specific feedback effect of frowning if the ratings when frowning were lower than the ratings in a neutral condition, since feedback from a frown is expected to decrease the ratings of pleasantness. Consequently, it would also be possible to conclude if the feedback effect of smiling is stronger with a positive emotion present (enhancing) or with a negative emotion present (attenuating), and if the feedback effect of frowning is stronger with a negative emotion present (enhancing) or with a positive emotion present (attenuating).

In the literature on facial feedback, both enhancing and attenuating modulation have been demonstrated, but there are no studies, as far as we know, that have compared the two in a manner that could investigate if feedback is more effective to enhance or attenuate emotions (Adelmann and Zajonc 1989 ; McIntosh 1996 ). Even though many of the first studies on facial feedback included neutral muscle conditions (e.g., Laird 1974 ; Rutledge and Hupka 1985 ), their focus was primarily on testing the main hypothesis of facial feedback, and studies were typically not designed or analyzed in a way that could answer the present questions at hand.

In a classical study by Strack et al. ( 1988 ), a neutral muscle condition was included in a design where participants were instructed to rate the funniness of cartoons with the help of a pen either held between the teeth, between the lips or in the non-dominant hand. Holding the pen between the teeth facilitated smiling since the participants needed to activate the zygomaticus major muscle in order to keep the lips from the pen. Holding the pen between the lips, on the other hand, inhibited smiling and also activated the orbicularis oris muscle, involved in the expression of anger (Ekman and Friesen 1975 ; Hjortsjö 1970 ). Holding the pen in the non-dominant hand was in this case the neutral muscle condition. The results demonstrated significant differences, where the funniness ratings were highest for participants who had been facilitated to smile, lowest for participants who had been inhibited to smile, and with the neutral condition perfectly placed in the middle. However, a recent attempt to directly replicate this experiment found no effects on the funniness ratings at all between the teeth and the lip positions of the pen (Wagenmakers et al. 2016 ). The absence of effects was unexpected since numerous studies over the years have used variations of the original method to successfully manipulate emotions, even if no direct replication had been done previously (Strack 2016 ).

Even though the present-day effectiveness of the original method must be questioned in light of the failed replication, the original study by Strack et al. ( 1988 ) still remains of interest for the present study since it demonstrated both enhancing and attenuating feedback effects. That is, with a positive emotion present, a smile enhanced the rated funniness while a mouth that expressed anger attenuated the rated funniness. In other words, the original results demonstrated that facial feedback was equally effective at enhancing and attenuating a positive emotion in this setting. However, it is necessary to note that holding a pen between the lips, and thereby activating the orbicularis oris , does not constitute a complete and valid expression of anger (Ekman and Friesen 1975 ; Hjortsjö 1970 ). An essential part of expressing anger is to frown and this action was not included in the study by Strack et al. ( 1988 ), and by extension not in the replication effort (Wagenmakers et al. 2016 ). Therefore, it is unclear if it is effective enough to compare these two muscle groups to each other, since it is possible that the rated funniness had been lower if the participants had also been instructed to frown to the cartoons. In summary, the original study by Strack et al. ( 1988 ) demonstrated that facial feedback had the ability to both enhance and attenuate present positive emotions, even though their relative effectiveness remained unknown. Note however that this result could not be replicated by Wagenmakers et al. ( 2016 ).

Soussignan ( 2002 ) adopted almost the same method as Strack et al. ( 1988 ), but with two separate intensity levels of smiling, one low and one high. The study had both positive and negative emotional stimuli, and the neutral muscle condition was to hold the pen in the mouth in a way that relaxed most muscles. The results showed that the high intensive smile group had significantly higher ratings than the other groups, but only to positive emotional stimuli. There were no feedback effects at all to negative stimuli. Furthermore, there was no difference in ratings between the group that had the pen between the lips and the neutral control group. That is, with a positive emotion present, a smile enhanced the ratings, but a mouth that expressed anger did not attenuate them, a result suggesting that facial feedback primarily enhances present positive emotions. However, since the study by Soussignan ( 2002 ) used a similar method to Strack et al. ( 1988 ), the same limitations apply to the angry mouth condition, which means that an attenuating effect remained a possibility even though no effects were found.

Thus, the present study addressed the question if facial feedback primarily enhances or attenuates emotions by introducing two neutral muscle control conditions to the voluntary facial action technique. In line with the cover story of measuring reaction times, the neutral conditions were to press a button on a handheld device and/or to do nothing. Consequently, the participants would ostensibly test their reaction times in two different phases: one facial phase where they would react with either a smile or a frown to pictures of happy and angry faces, and one neutral phase where they would react by either pressing a button or doing nothing to the same set of pictures. By comparing the effects of the smile and the frown conditions with the neutral conditions, it would then be possible to evaluate (1) if facial feedback primarily enhances emotions, a situation where smiles enhance positive emotions and frowns enhance negative emotions, or (2) if feedback primarily attenuates emotions, a situation where smiles attenuate negative emotions and frowns attenuate positive emotions, or (3) if feedback both enhances and attenuates emotions.

The second question, if smiling or frowning produces a stronger feedback effect than the other, is closely related to the first question. The happy smile and the angry frown are powerful emotional expressions and it is well established that both induce facial feedback. However, it remains unknown if smiling and frowning are equally effective when it comes to modulating our present emotions. In Experiment 1, as well as in Dimberg and Söderkvist ( 2011 ), a general feedback effect was calculated by the difference in ratings between smiling and frowning conditions. With such procedure however it is not possible to tell if the feedback comes from smiling, from frowning, or from both. Thus, as in the above first question, a neutral muscle condition is necessary in order to discern the specific feedback effects of smiling and frowning. Consequently, the present study addressed this question with the help of the two neutral muscle conditions described above.

The third question, if positive or negative emotions are easier to modulate with feedback, was addressed by Dimberg and Söderkvist ( 2011 ) who found no difference between positive and negative emotions in regards to feedback effects. However, in the study by Soussignan ( 2002 ) mentioned above, significant feedback effects were only found during positive stimuli, with no effects during negative stimuli. Moreover, and in contrast, a study by Davis et al. ( 2009 ) found feedback effects during negative stimuli, but not during positive, when they compared inhibited and non-inhibited facial reactions to emotional stimuli. The aim of the present study was to further investigate this question. Note that this third question is intimately related to the first and second questions in the sense that some specific outcomes of those two questions could provide the answer to this question. For example, if facial feedback primarily enhances present emotions, at the same time as feedback effects from smiling are stronger, then it follows that facial feedback modulates positive emotions more effectively. Considering that previous studies have demonstrated inconsistent results, the present study had no specific expectations regarding the outcome.

Sixty-four persons (mean age = 23.75, SD  = 3.98), balanced by gender, participated in the experiment and most of them were students at Uppsala University. No one that studied psychology or previously had participated in similar studies was included. Five persons were excluded from the study due to not following the instructions correctly, and five persons were excluded due to realizing the true purpose of the study, and they were all replaced by other participants. As compensation, participants were given a movie voucher at a value of 100 SEK (approximately 10 USD).

The two independent variables of the experiment were muscle (smile vs. frown vs. button vs. nothing) and stimulus (happy vs. angry), and it thus had a within subject 4 × 2 factorial design. The dependent variable was the ratings of how pleasant/unpleasant the participants experienced the stimuli.

Similar to the procedure in Experiment 1, the participants were individually tested in a small room inside the psychophysiological laboratory, where they were seated in front of a computer monitor. In the facial phase, where they reacted with either a smile or a frown, electrodes to ostensibly measure EMG were attached over the zygomatic major (smiling) and the corrugator supercilii (frowning) muscle regions. In the neutral phase, where they reacted by either pressing a button or doing nothing, a handheld device was used to seemingly register their reaction times.

The stimulus material consisted of the same six pictures of happy faces and six pictures of angry faces (Ekman and Friesen 1976 ) that were successfully used in Experiment 1, and another reason for using these stimuli was once again to maintain consistency with the experiments in Dimberg and Söderkvist ( 2011 ). The pictures were presented in the form of PowerPoint slideshows, where each picture was displayed for 8 s and then followed by a black screen for 20, 25 or 30 s. Participants rated on two different scales of 0–100 ( not at all to very much ) how pleasant and unpleasant they experienced the pictures. For each picture, a sheet containing both scales was used.

As with the cover story in Experiment 1, participants were told that the aim of the experiment was to measure the reaction times in different muscles. Half of the participants were randomized to begin with the facial phase, where they had electrodes attached on their facial muscles and were instructed to react with either their smile or their frown muscles. After a short break, they continued with the neutral phase where they were introduced to the handheld device and were instructed to react either by pressing a button on the device or by doing nothing. The other half of participants began with the neutral phase, and then continued with the facial phase after a short break. Two different neutral conditions were needed in order for the neutral phase to be as similar as possible to the facial phase, and to seem a plausible test of reaction times.

Within each phase, the participants performed two different rating tasks. In each of these tasks they were exposed to all the six happy and six angry pictures, in different orders, and were instructed to rate how pleasant and unpleasant they experienced them. In the facial phase, one task was to react by elevating the cheeks as fast as possible if the displayed picture was of a happy person, and to contract the eyebrows as fast as possible if the picture was of an angry person. The other task was accompanied by the reversed instruction, that is, to elevate the cheeks if the picture was of an angry person and to contract the eyebrows if the picture was of a happy person. The order of tasks was balanced amongst the participants. In the neutral phase, one task was to react by pressing a button on the handheld device as fast as possible if the displayed picture was of a happy person, and to do nothing if the picture was of an angry person. The other task was with the reversed instructions, that is, to press a button on the handheld device if the picture was of an angry person and to do nothing if the picture was of a happy person. After the experiment, participants were interviewed in order to determine if they had followed the instructions correctly and if they at any time had realized the true purpose of the experiment.

As a result of the procedure, each participant had rated how pleasant and unpleasant they found the six happy and six angry pictures during the four different muscle conditions: smiling, frowning, pressing a button, and doing nothing. Before further analysis, similar to the procedure in Experiment 1 and earlier studies (Dimberg and Söderkvist 2011 ), we compared the pleasantness and unpleasantness ratings to see if there were any major differences between them. When we correlated the two ratings within each participant the median correlation coefficient for all participants was −0.80 ( SIQR  = 0.13), and the results provided by the pleasantness and the unpleasantness ratings pointed in the same direction. As in Experiment 1, we therefore merged the two ratings to a single rating score by subtracting the unpleasantness ratings from the pleasantness ratings, resulting in the same bipolar scale measuring from −100 to 100 used in Experiment 1. Moreover, the ratings of the two neutral muscle conditions were compared and it was found that there was no significant difference between them, t (63) = 0.01, ns , and also that they had a strong positive correlation, r  = 0.90, p  < 0.05. Thus, in order to simplify the analysis, the two ratings were merged into a single neutral rating.

Because the purpose of the study was to measure and compare the magnitude of the facial feedback effects of smiling and frowning in enhancing and attenuating modulation respectively a new dependent variable specifically representing each muscle action was created. The specific facial feedback effect (SFFE) scores were calculated by comparing the ratings when smiling and the ratings when frowning to the ratings in the neutral condition. That is, the SFFE score for smiling was calculated by subtracting the ratings in the neutral condition from the ratings when smiling . This procedure created a score that had a positive value if the ratings when smiling were higher than the ratings in the neutral condition. The SFFE score for frowning , on the other hand, was calculated by subtracting the ratings when frowning from the ratings in the neutral condition , which created a score that had a positive value if the ratings when frowning were lower than the ratings in the neutral condition. To summarize, these calculations transformed the ratings into a score that expresses absolute values of feedback effect . That is, with this score a positive sign indicates a feedback effect for both smiling and frowning independent of if the facial action originally increased or decreased the rated pleasantness. As a first step in creating these new dependent variables, the SFFE scores for smiling and frowning were calculated for each rated picture. Secondly, for each participant the scores of the six happy and the six angry pictures were collapsed into one mean for the happy and one mean for the angry pictures, for smiling and frowning respectively. This resulted in four mean SFFE scores for each participant, that is, one score for smiling and one score for frowning during positive emotions (to happy faces), and one for smiling and one for frowning during negative emotions (to angry faces).

The three questions in the present study were first, if facial feedback primarily enhances or attenuates emotions, second, if smiling or frowning produces a stronger feedback effect than the other, and finally if positive or negative emotions are easier to modulate with feedback. In the analysis, these three questions were evaluated using three different factors, one for modulation (enhancing vs. attenuating), one for muscle (smiling vs. frowning), and one for emotion (positive vs. negative). Note however that due to inherent properties of these factors there exists an inevitable interdependency between them. For example, smiling as an enhancing modulation is only possible during a positive emotion and never during a negative emotion, while frowning as an enhancing modulation is only possible during a negative emotion. Similarly, smiling during a positive emotion is per definition an enhancing modulation and cannot be attenuating, while frowning during a positive emotion is always an attenuating modulation. That is, the muscle factor is intertwined and linked together with the modulation and emotion factors in these cases, which illustrates the fact that the conditions within each of the three factors by nature are confounded with each other. Thus, due to this interdependency between the factors it was not possible to simultaneously evaluate the effects in a three-factorial analysis of variance. Instead the three questions were addressed in three separate one-factorial analyses. That is, in each analysis, the four SFFE scores for each participant were collapsed in a different way into two mean scores. First, the mean score for enhancing modulation consisted of the SFFE scores for when participants smiled during positive emotions and frowned during negative emotions, while the mean score for attenuating modulation consisted of the SFFE scores for when participants smiled during negative emotions and frowned during positive emotions. Second, the mean score for smiling consisted of the SFFE scores for when they smiled during positive and during negative emotions, while the mean score for frowning consisted of the SFFE scores for when they frowned during positive and during negative emotions. Finally, the mean score for positive emotions consisted of the SFFE scores for when they smiled and frowned during positive emotions, while the mean score for negative emotions consisted of the SFFE scores for when they smiled and frowned during negative emotions.

As explained in the method section above, the three main questions were evaluated in three separate one-factorial ANOVAs, with the data reconstructed for each analysis. The mean SFFE scores for the two conditions of each of the three factors are presented in Fig.  3 .

Mean specific facial feedback effect (SFFE) scores for (left) enhancing and attenuating modulation, (middle) smiling and frowning, and (right) during positive and negative emotions, with error bars representing one standard error

Enhancing Versus Attenuating

As detailed above, in order to compare the feedback effects between enhancing and attenuating modulation, the four mean SFFE scores of each participant were collapsed into two mean scores, one for enhancing and one for attenuating modulation. To analyze the resulting data, a one-way within subject ANOVA with the factor modulation (enhancing vs. attenuating) was performed. The result demonstrated a significant effect of modulation, F (1,63) = 4.15, MS e  = 210.58, p  < 0.05, η p 2  = 0.06, and as seen in the left panel of Fig.  3 this effect was due to higher SFFE scores in attenuating [ M  = 4.29, SD  = 12.45, 95% CI (1.18, 7.40)] as compared to enhancing modulation [ M  = −0.93, SD  = 9.68, 95% CI (−3.35, 1.48)]. In order to verify that the attenuating effect per se was significant, we performed additional one-sample t -tests to determine if the scores differed from zero. The tests demonstrated that the attenuating score, t (63) = 2.36, p  < 0.05, revealed a significant feedback effect, while the enhancing score did not differ from zero, t  < 1.

Smile Versus Frown

To compare the feedback effects between smiling and frowning, the four mean SFFE scores of each participant were collapsed into two mean scores, one for smiling and one for frowning. The data were analyzed with a one-way within subject ANOVA with the factor muscle (smile vs. frown). The result showed no effect of muscle, F (1,63) = 0.31, MS e  = 153.58, ns , η p 2  = 0.01, and as seen in the middle panel of Fig.  3 there was no major difference in SFFE scores between when participants smiled [ M  = 1.07, SD  = 10.11, 95% CI (−1.46, 3.59)] and frowned [ M  = 2.29, SD  = 9.47, 95% CI (−0.08, 4.65)].

Positive Versus Negative

To compare the feedback effects during positive emotions with the effects during negative emotions, the four mean SFFE scores of each participant were collapsed into a final set of two mean scores, one for positive emotions and one for negative emotions. To analyze the data, a one-way within subject ANOVA with the factor emotion (positive vs. negative) was performed. As can be seen in the right panel of Fig.  3 there was a non-significant trend towards slightly larger scores to positive [ M  = 2.77, SD  = 6.60, 95% CI (1.12, 4.42)] than negative emotions [ M  = 0.59, SD  = 6.43, 95% CI (−1.02, 2.19)], F (1,63) = 3.27, MS e  = 46.65, p  < 0.10, η p 2  = 0.05.

Our findings suggest that facial feedback primarily affects present emotions when the facial action is incongruent with the emotional valence. In fact, there were no indications of feedback effects at all when the facial action and the emotion were congruent. This absence of a feedback effect in enhancing modulation is unexpected considering that several previous studies have demonstrated enhancing effects (e.g., Soussignan 2002 ; Strack et al. 1988 ). One explanation for the diverging results could be that the present Experiment 2 was specifically designed to evaluate possible differences between enhancing and attenuating modulation. That is, in contrast to previous studies, the present study did not only let participants both smile and frown to positive and to negative stimuli, but instead, data were transformed and analyzed in a way making it possible to combine enhancing smiles and frowns, and combining attenuating smiles and frowns, in order to evaluate the overall enhancing and attenuating effects per se . It seems that this approach was effective in revealing attenuating effects. Still, since the attenuating effect in the present study was unexpected, it would be valuable with more research on the subject before drawing definitive conclusions. In light of this we realized that the data in the present Experiment 1 in fact could be transformed and reanalyzed in a way that would allow us to further test the main questions of Experiment 2. This reanalysis is presented in the Reanalysis of Experiment 1 section below.

Regarding the two other questions of the present study, no significant differences were found. That is, the acts of smiling and frowning appeared to produce feedback effects to a similar degree. Moreover, there was no significant difference in feedback effects during positive and negative emotions, which is in line with earlier results (Dimberg and Söderkvist 2011 ).

Finally, five participants in Experiment 2 realized the true purpose of the study, as compared to Experiment 1 in which no participants realized the true purpose. However, in the interviews after the experiment this was explained by the fact that these participants had heard about and were familiar with the facial feedback theory.

Reanalysis of Experiment 1

If the significant difference in feedback effects between enhancing and attenuating modulation in Experiment 2 represents a genuine attenuating effect it should be possible to replicate this effect in other experiments where the SFFE scores for smiling and frowning can be calculated for enhancing and attenuating modulation. As it happens, and mentioned above, Experiment 1 in the present study can be considered such an experiment, which presents us with an immediate opportunity to test if the unexpected attenuating effect in Experiment 2 can be replicated.

That is, in the results of Experiment 1 it was demonstrated that there were no effects at all of facial feedback after 5 min and after 1 day. Consequently, these later conditions, and especially the condition after 1 day, could be assumed to reflect a more neutral muscle condition where participants were not influenced by voluntary performed facial actions. As detailed in Experiment 2, a neutral muscle condition makes it possible to discern the specific feedback effects of smiling and frowning, which in turn enable us to address the present questions.

Following the same procedure for transformations as in Experiment 2, and using the ratings after 1 day as the neutral muscle condition, new dependent variables were created in the form of SFFE scores for smiling and frowning. As a result of the transformations, four mean SFFE scores for each participant were created, that is, one score for smiling and one score for frowning during positive emotions, and one for smiling and one for frowning during negative emotions. Similar to Experiment 2, the three questions were evaluated by help of three different factors, one for modulation (enhancing vs. attenuating), one for muscle (smiling vs. frowning), and one for emotion (positive vs. negative), and the factors were analyzed in three separate one-factorial ANOVAs, with the data reconstructed for each analysis.

To compare the feedback effects between enhancing and attenuating modulation, the four mean SFFE scores of each participant were, similarly as in Experiment 2, collapsed into two mean scores, one for enhancing and one for attenuating modulation. The result demonstrated a significant effect of modulation, F (1,31) = 10.92, MS e  = 255.49, p  < 0.05, η p 2  = 0.26, with higher SFFE scores in attenuating [ M  = 10.31, SD  = 16.40, 95% CI (4.40, 16.23)] as compared to enhancing modulation [ M  = −2.89, SD  = 13.13, 95% CI (−7.63, 1.84)]. In order to verify that the attenuating effect per se was significant, we performed additional one-sample t -tests to determine if the scores differed from zero. The tests demonstrated that the attenuating score, t (31) = 3.65, p  < 0.05, reflected a significant feedback effect, while the enhancing score did not differ from zero, t  < 1.05.

Regarding the two other questions, there was no significant difference in SFFE scores between when the participants smiled [ M  = 6.60, SD  = 11.92, 95% CI (2.30, 10.90)] and frowned [ M  = 0.82, SD  = 13.20, 95% CI (−3.94, 5.58)], F (1,31) = 4.09, MS e  = 130.53, p  < 0.10, η p 2  = 0.12, and there was no significant difference during positive emotions [ M  = 4.75, SD  = 12.00, 95% CI (0.42, 9.07)] or negative emotions [ M  = 2.67, SD  = 14.44, 95% CI (−2.54, 7.88)], F (1,31) = 0.42, MS e  = 166.57, ns , η p 2  = 0.01.

The results of this reanalysis demonstrate that facial feedback almost exclusively attenuated emotions, with quite a large effect size. This replicates the results of Experiment 2 and further confirms that facial feedback primarily affects ongoing emotions when the facial action is incongruent to the emotion. Just as in Experiment 2 there were no indications of any feedback effects when the facial action was congruent to the emotion.

General Discussion

The purpose of the present study was to further detail the description of how facial feedback effects occur. In two experiments, four different questions were addressed. In Experiment 1, it was demonstrated that the effect of facial feedback primarily occurs at the same time when the facial action is performed. In Experiment 2 it was demonstrated that facial feedback primarily attenuates ongoing emotions, and the same pattern was evident in the Reanalysis of Experiment 1. Notably, there seems to be no systematic difference in feedback effects between smiles and frowns, or between positive and negative emotions.

Regarding the finding that facial feedback primarily attenuates present emotions there are a number of studies in the literature on facial reactions to emotional stimuli (e.g., Dimberg et al. 2000 , 2002 ) that provide important insights. First, when people are exposed to emotional facial stimuli they spontaneously and unconsciously react with their corresponding facial muscles. These reactions are automatic, very quick, and not always visible to the eye, but can be detected with facial electromyographic equipment. Second, these automatic reactions affect voluntary facial actions by facilitating congruent actions. For example, people that are exposed to a positive stimulus react quickly and automatically with their smile muscles even before they are consciously aware of the valence of the stimulus. Then, if they are instructed to voluntarily react with a smile and/or a frown to the positive stimulus, the voluntary smile will be formed faster and be more pronounced than the voluntary frown that requires more time and is less pronounced. The same is true for a negative stimulus, which facilitates a faster and stronger voluntary frown, as compared to a voluntary smile. Therefore, since these automatic reactions are present even when people are instructed to not move their facial muscles, one could question if it is possible at all to expose participants to a truly neutral muscle condition in an experiment that uses emotional facial stimuli, such as in the present study. That is, when we compared the difference in feedback effects between voluntary facial actions and neutral muscle conditions, automatic facial reactions similar to those described above were probably initially present in all conditions. It is therefore interesting that even though a congruent facial action is faster and stronger, as compared to an incongruent action (Dimberg et al. 2002 ), it was the latter actions that primarily modulated ongoing emotional experiences.

In the facial feedback literature there are no obvious explanations for the attenuating effect found in the present study. One potential explanation could perhaps be found in the above mentioned field of automatic facial reactions to emotional stimuli (e.g., Dimberg et al. 2000 , 2002 ), where recent studies (e.g., Wood et al. 2016a , b ) suggest that incongruent facial actions can disrupt the automatic mimicry of facial stimuli and that this lack of mimicry might lower the general impact of the stimulus. If this is the case it could offer an explanation for the attenuating effect of incongruent facial actions, but it would not explain the absence of an enhancing effect from congruent actions.

Another potential explanation, based on the idea that emotions are governed by different affect programs (Tomkins 1962 ), could be that one function of emotional facial actions is to initiate corresponding affect programs. That is, if no emotions were present a specific facial action would initiate a corresponding affect program and thus initiate an emotional experience. With an emotion already present, however, a congruent facial action would probably have no or little effect since the corresponding affect program already would have been initiated by, for example, an emotional stimulus. An incongruent facial action, on the other hand, would initiate a conflicting affect program, which might reduce the activation of the originally initiated program. Admittedly, while this explanation would account for the present results it remains hard to test empirically.

Finally, there is also the possibility that the present attenuation effect is the result of a ceiling effect in the intensity of the induced emotions. In such a case, the presented stimuli evoked the emotions to a maximum level and simply left no room for further enhancement, but plenty of room for an incongruent action to cause attenuation. This is unlikely though since the present stimuli, pictures of facial expressions, are not expected to induce strong emotions (e.g., Wangelin et al. 2012 ). A suggestion for future research could be to investigate if there is any difference in general feedback effects, and between enhancing and attenuating modulation, when stimuli of different intensity levels are used. It would be interesting to see if the feedback effects become stronger with stimuli that induce very strong emotions, or if the effects even become weaker (e.g., Baumeister et al. 2016 ). Moreover, in order to eliminate the possibility of a ceiling effect it could also be interesting to use stimuli that is expected to induce weaker emotions, for example pictures of mildly happy and mildly angry persons, which would allow congruent actions to enhance the evoked emotions in case of a ceiling effect.

The effort by Wagenmakers et al. ( 2016 ) to replicate the well-known feedback study by Strack et al. ( 1988 ) is the largest and most ambitious replication study of facial feedback to this day. As discussed earlier, they could not replicate the original results which questions the original study, but it should also be noted that some of the methodological choices made in the replication study have been questioned by Strack ( 2016 ). It is also important to note, as Wagenmakers et al. ( 2016 ) do as well, that the failed replication does not mean a general refutation of the facial feedback hypothesis. Indeed, as the results of the present study also suggest, facial feedback effects have been demonstrated in numerous studies that cover a wide range of different methods. For this reason there is a need for more replication efforts and a continued systematic evaluation of the facial feedback field, and the scope and ambition of the effort by Wagenmakers et al. ( 2016 ) has certainly set the standard for future such studies.

Finally, the present results may have potential implications for the clinical studies that investigate the effect of Botox in the treatment of depression (e.g., Finzi and Rosenthal 2014 ; Magid et al. 2014 ; Wollmer et al. 2012 , 2014 ). If, as we found in the present study, facial feedback primarily attenuates present emotions, and further, if constant frowning contributes to the depressed state, then this frowning may act to specifically attenuate positive emotions. The paralysis of frown muscles with Botox could then block this attenuation, which may result in a general recovery of positive emotions, and consequently this effect could be one basis for the documented improvements of the patients. Future research could study how Botox affects the day-to-day emotional life of depressed patients in order to find out in which situations the improvements are experienced most frequently.

The second question of Experiment 2 was if smiling or frowning produces a stronger feedback effect than the other. The results indicated that there were no significant differences in effect between them, which suggest that a smile and a frown produce feedback effects that are roughly equal in strength. That is, in relation to the specific emotional stimuli used in these experiments, they produced equally strong feedback. One interesting question is if the smile and the frown actions might differ in how easy they are to activate and/or differ in the strength of the activation itself. This question has earlier been addressed in studies that also used the voluntary facial action technique (for a discussion see Dimberg and Söderkvist 2011 ). In that study it was noted that the strength of the muscle activity for smiles and frowns did not differ and further that the activity levels over repeated trials were not differently affected by for example fatigue. This suggests that the two different facial actions should be comparable to each other. Moreover, the stimulus material consisted of clearly positive and clearly negative pictures, and it is possible that the relative strength between smiling and frowning could be different with another type of stimuli. Future research could therefore use more neutral or ambiguous stimuli to further investigate this question, and if completely neutral stimuli were used it would allow for a comparison of the initiating ability of facial feedback between smiling and frowning.

The third question of Experiment 2 was if positive or negative emotions are easier to modulate with feedback. This question has been addressed in both Dimberg and Söderkvist ( 2011 ) and in the present study, but none of these studies have found any significant differences between positive and negative emotions. There simply seems to be no difference between them in this regard.

In the present study it was assumed that the participants followed the instructions correctly. They were not monitored during the experiments, but we have no indications that they found the instructions difficult to follow. The control questions afterward aimed to find out if they had realized the true purpose, but also aimed to verify that they had followed the instructions. The few participants that did not follow the instructions correctly were easily identified. Note also that if a participant should not follow the muscle instructions correctly it would in fact work against the facial feedback hypothesis and consequently reduce the general level of facial feedback effect in the present experiments.

As mentioned above, the purpose of the present study was to further detail the description of how facial feedback occurs. The facial feedback hypothesis, as discussed by Adelmann and Zajonc ( 1989 ), defines facial feedback as facial muscle actions that have the ability to modulate present emotions or, in the absence of emotions, to initiate emotions. Collectively then, we would like to add that the effect of facial feedback mainly occurs at the time when the facial action is performed, and that facial feedback primarily attenuates ongoing emotional experiences.

Acknowledgements

This project was supported by a grant to Ulf Dimberg from the Faculty of Social Sciences at Uppsala University. We thank Mats Fredrikson and two anonymous reviewers for valuable comments on the article.

Compliance with Ethical standards

Conflict of interest.

The authors declare that they have no conflict of interest.

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors.

Informed Consent

Informed consent was obtained from all individual participants included in the study.

• Adelmann PK, Zajonc RB. Facial efference and the experience of emotion. Annual Review of Psychology. 1989; 40 :249–280. doi: 10.1146/annurev.ps.40.020189.001341. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Baumeister JC, Papa G, Foroni F. Deeper than skin deep–The effect of botulinum toxin-A on emotion processing. Toxicon. 2016; 118 :86–90. doi: 10.1016/j.toxicon.2016.04.044. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Buck R. Nonverbal behavior and the theory of emotion: The facial feedback hypothesis. Journal of Personality and Social Psychology. 1980; 38 :811–824. doi: 10.1037/0022-3514.38.5.811. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Darwin C. The expression of emotions in man and animals. London: Murray; 1872. [ Google Scholar ]
• Davis JI, Senghas A, Ochsner KN. How does facial feedback modulate emotional experience? Journal of Research in Personality. 2009; 43 (5):822–829. doi: 10.1016/j.jrp.2009.06.005. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dimberg U. Facial reactions to facial expressions. Psychophysiology. 1982; 19 :643–647. doi: 10.1111/j.1469-8986.1982.tb02516.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dimberg U. Facial reactions to fear-relevant stimuli for subjects high and low in specific fear. Scandinavian Journal of Psychology. 1990; 31 :65–69. doi: 10.1111/j.1467-9450.1990.tb00803.x. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dimberg U, Hansson G, Thunberg M. Fear of snakes and facial reactions: A case of rapid emotional responding. Scandinavian Journal of Psychology. 1998; 39 :75–80. doi: 10.1111/1467-9450.00059. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dimberg U, Söderkvist S. The voluntary facial action technique: A method to test the facial feedback hypothesis. Journal of Nonverbal Behavior. 2011; 35 :17–33. doi: 10.1007/s10919-010-0098-6. [ CrossRef ] [ Google Scholar ]
• Dimberg U, Thunberg M, Elmehed K. Unconscious facial reactions to emotional facial expressions. Psychological Science. 2000; 11 :86–89. doi: 10.1111/1467-9280.00221. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Dimberg U, Thunberg M, Grunedal S. Facial reactions to emotional stimuli: Automatically controlled emotional responses. Cognition and Emotion. 2002; 16 :449–471. doi: 10.1080/02699930143000356. [ CrossRef ] [ Google Scholar ]
• Ekman P. Cross-cultural studies of facial expressions. In: Ekman P, editor. Darwin and facial expression. New York: Academic Press; 1973. [ Google Scholar ]
• Ekman P, Friesen WV. Unmasking the face: A guide to recognizing emotions from facial cues. Englewood Cliffs, NJ: Prentice Hall; 1975. [ Google Scholar ]
• Ekman P, Friesen WV. Pictures of facial affect. Palo Alto: Consulting Psychologists Press; 1976. [ Google Scholar ]
• Finzi E, Rosenthal NE. Treatment of depression with onabotulinumtoxinA: A randomized, double-blind, placebo controlled trial. Journal of Psychiatric Research. 2014; 52 :1–6. doi: 10.1016/j.jpsychires.2013.11.006. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Flack WF, Laird JD, Cavallaro LA. Separate and combined effects of facial expressions and bodily postures on emotional feelings. European Journal of Social Psychology. 1999; 29 :203–217. doi: 10.1002/(SICI)1099-0992(199903/05)29:2/3<203::AID-EJSP924>3.0.CO;2-8. [ CrossRef ] [ Google Scholar ]
• Gellhorn E. Motion and emotion: The role of proprioception in the physiology and pathology of the emotions. Psychological Review. 1964; 71 :457–472. doi: 10.1037/h0039834. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hennenlotter A, Dresel C, Castrop F, Ceballos Baumann AO, Wohlschläger AM, Haslinger B. The link between facial feedback and neural activity within central circuitries of emotion—new insights from Botulinum Toxin—induced denervation of frown muscles. Cerebral Cortex. 2009; 19 :537–542. doi: 10.1093/cercor/bhn104. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Hjortsjö CH. Man’s face and mimic language. Malmö: Nordens Boktryckeri; 1970. [ Google Scholar ]
• Hofmann W, De Houwer J, Perugini M, Baeyens F, Crombez G. Evaluative conditioning in humans: a meta-analysis. Psychological Bulletin. 2010; 136 (3):390. doi: 10.1037/a0018916. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Izard CE. The face of emotion. New York: Plenum; 1971. [ Google Scholar ]
• James W. What is emotion? Mind. 1884; 19 :188–205. doi: 10.1093/mind/os-IX.34.188. [ CrossRef ] [ Google Scholar ]
• Kim MJ, Neta M, Davis FC, Ruberry EJ, Dinescu D, Heatherton TF, Stotland MA, Whalen PJ. Botulinum toxin-induced facial muscle paralysis affects amygdala responses to the perception of emotional expressions: Preliminary findings from an ABA design. Biology of Mood and Anxiety Disorders. 2014; 4 (1):11. doi: 10.1186/2045-5380-4-11. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Laird JD. Self-attribution of emotion: The effects of expressive behavior on the quality of emotional experience. Journal of Personality and Social Psychology. 1974; 29 (4):475–486. doi: 10.1037/h0036125. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Lewis MB. Exploring the positive and negative implications of facial feedback. Emotion. 2012; 12 (4):852–859. doi: 10.1037/a0029275. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Magid M, Reichenberg JS, Poth PE, Robertson HT, LaViolette AK, Kruger TH, Wollmer MA. Treatment of major depressive disorder using botulinum toxin A: A 24-week randomized, double-blind, placebo-controlled study. The Journal of Clinical Psychiatry. 2014; 75 (8):837–844. doi: 10.4088/JCP.13m08845. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• McIntosh DN. Facial feedback hypothesis: Evidence, implications, and directions. Motivation and Emotion. 1996; 20 :121–147. doi: 10.1007/BF02253868. [ CrossRef ] [ Google Scholar ]
• Mori K, Mori H. Examination of the passive facial feedback hypothesis using an implicit measure: With a furrowed brow, neutral objects with pleasant primes look less appealing. Perceptual and Motor Skills. 2010; 111 (3):785–789. doi: 10.2466/02.07.24.PMS.111.6.785-789. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Mori H, Mori K. An implicit assessment of the effect of artificial cheek raising: When your face smiles, the world looks nicer. Perceptual and Motor Skills. 2013; 116 (2):466–471. doi: 10.2466/24.50.PMS.116.2.466-471. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Niedenthal PM. Embodying emotion. Science. 2007; 316 :1002–1005. doi: 10.1126/science.1136930. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Niedenthal PM, Mermillod M, Maringer M, Hess U. The Simulation of Smiles (SIMS) model: Embodied simulation and the meaning of facial expression. Behavioral and Brain Sciences. 2010; 33 (6):417–433. doi: 10.1017/S0140525X10000865. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Niedenthal PM, Mondillon L, Winkielman P, Vermeulen N. Embodiment of emotion concepts. Journal of Personality and Social Psychology. 2009; 96 (6):1120–1136. doi: 10.1037/a0015574. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Rutledge LL, Hupka RB. The facial feedback hypothesis: Methodological concerns and new supporting evidence. Motivation and Emotion. 1985; 9 (3):219–240. doi: 10.1007/BF00991829. [ CrossRef ] [ Google Scholar ]
• Soussignan R. Duchenne smile, emotional experience, and autonomic reactivity: A test of the facial feedback hypothesis. Emotion. 2002; 2 :52–74. doi: 10.1037/1528-3542.2.1.52. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Strack, F. (2016). Reflection on the smiling registered replication report.  Perspectives on Psychological Science , 1745691616674460. [ PubMed ]
• Strack F, Martin FF, Stepper S. Inhibiting and facilitating conditions of the human smile: A non-obtrusive test of the facial feedback hypothesis. Journal of Personality and Social Psychology. 1988; 54 :768–777. doi: 10.1037/0022-3514.54.5.768. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Tomkins SS. Affect, imagery, consciousness: Vol. 1. The positive affects. New York: Springer; 1962. [ Google Scholar ]
• Tomkins SS. Affect as amplification: some modifications in theory. In: Plutchik R, Kellerman H, editors. Emotion: Theory, research and experience. New York: Academic Press; 1980. [ Google Scholar ]
• Wagenmakers EJ, Beek T, Dijkhoff L, Gronau QF, Acosta A, Adams RB, Jr, Bulnes LC. Registered replication report: Strack, Martin, & Stepper (1988) Perspectives on Psychological Science. 2016; 11 (6):917–928. doi: 10.1177/1745691616674458. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wangelin BC, Bradley MM, Kastner A, Lang PJ. Affective engagement for facial expressions and emotional scenes: the influence of social anxiety. Biological Psychology. 2012; 91 (1):103–110. doi: 10.1016/j.biopsycho.2012.05.002. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wollmer MA, de Boer C, Kalak N, Beck J, Gotz T, Schmidt T, Kruger THC. Facing depression with botulinum toxin: A randomized controlled trial. Journal of Psychiatric Research. 2012; 46 (5):574–581. doi: 10.1016/j.jpsychires.2012.01.027. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wollmer MA, Kalak N, Jung S, de Boer C, Magid M, Reichenberg JS, Kruger THC. Agitation predicts response of depression to botulinum toxin treatment in a randomized controlled trial. Frontiers in Psychiatry. 2014; 5 :36. doi: 10.3389/fpsyt.2014.00036. [ PMC free article ] [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wood A, Lupyan G, Sherrin S, Niedenthal P. Altering sensorimotor feedback disrupts visual discrimination of facial expressions. Psychonomic Bulletin & Review. 2016; 23 (4):1150–1156. doi: 10.3758/s13423-015-0974-5. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Wood A, Rychlowska M, Korb S, Niedenthal P. Fashioning the face: Sensorimotor simulation contributes to facial expression recognition. Trends in Cognitive Sciences. 2016; 20 (3):227–240. doi: 10.1016/j.tics.2015.12.010. [ PubMed ] [ CrossRef ] [ Google Scholar ]
• Zajonc RB, Murphy ST, Inglehart M. Feeling and Facial Efference: Implications of the Vascular Theory of Emotion. Psychological Review. 1989; 96 (3):395–416. doi: 10.1037/0033-295X.96.3.395. [ PubMed ] [ CrossRef ] [ Google Scholar ]

What Is The Facial Feedback Hypothesis And Does It Work?

Emotions are a basic part of the human experience, and expressing those emotions appropriately can be a part of improving one’s mental health. We can show our emotions in many ways, but one of the quickest and most common ways is through facial expressions. Facial expressions can do more than show others how we feel. Scientists have proposed the facial feedback hypothesis, which suggests that changing our facial expressions can also change our emotions.

What is the facial feedback hypothesis?

The  facial feedback hypothesis states that our facial expressions affect our emotions. If the facial-feedback hypothesis is correct, then not only do we smile when we feel happy, but smiling can make us feel happy, too. According to this hypothesis, in these cases, it is the act of smiling that produces a happy feeling. The same might hold true for other emotions as well.

Scientists have been interested in the idea of a facial-feedback hypothesis since the 1800s. In the 1840s, William James presented the idea that awareness of your bodily experiences is the basis of emotion. Thus, if you know that certain facial expressions are the ones you associate with being sad, you may experience a feeling of sadness.

Darwin investigated the way animals used facial expressions and suggested the idea of facial feedback in the 1870s. Through the latter half of the 1900s, the topic of facial feedback became popular again. Since then, many different studies have been done to test this hypothesis.

Types of facial expressions

What types of facial expressions can produce the emotions we feel? The scientific community is still debating how the facial feedback hypothesis might work with different expressions. One thing that seems certain, though, is that a smile is connected to the production of a happy emotion, while a frown is connected to a feeling of sadness.

Basic emotions

Degrees of emotions.

Along with the type of emotion we feel, we can also show the degree of that emotion through our facial expressions. For example, you may be slightly angry and express that emotion with a slight frown and furrowed eyebrows. If you're furious, though, your expressions will likely be much more distinctive.

Complex emotions

Often, we may feel combinations of emotions. Emotions aren't always pure or easily defined. Some common complex emotions are joyful love, prideful anger, and ambivalence. Complex emotions can be expressed with subtle variations of the usual facial expressions.

Duchenne vs. non-Duchenne smiles

A Duchenne smile is a  genuine smile , while a non-Duchenne smile is a fake smile. Although these two types of smiles are differentiated by whether the smile expresses an honest emotion you're feeling, you can make either expression whether you're already feeling happy or not.

In the non-Duchenne smile, you simply raise the corners of your mouth. It's what you might do when someone is going to take a photograph of you. They say, “Say cheese,” and you comply with a non-Duchenne smile.

The Duchenne smile starts with that same facial contraction, but it also involves raising your cheeks and squeezing your eyes. Your involuntary muscles do the extra work. So, how can you produce a Duchenne smile if you can't actively control those muscles? If we know what a Duchenne smile looks like, most of us can produce the same expression.

In a 2019 study, scientists carried out two experiments around Duchenne and non-Duchenne smiling. One evaluated how ostracism influenced the expression of emotion in a social environment, while the second replicated the results of the first experiment but focused particularly on smiling and self-reported emotion. In the first experiment, the participants who had higher frequencies of Duchenne smiling even during exclusion from the conversation self-reported higher rates of happiness. In the second, the relationship between non-Duchenne smiling and self-reported happiness was negative. However, certain participants were able to control their own emotional experience even while being ostracized, which led to an unexpected up-regulation of positive emotions .

Individual and cultural differences

Although all humans have many of the same basic facial expressions, some expressions may be unique to a specific individual or culture. So, if you know the person or culture well, it may be easier to understand what someone is expressing through facial expressions.

How the facial muscles express emotions

We often express emotions in our bodies, especially by using our facial muscles in specific ways. Why do we do it? How do we know how to hold our faces to show our emotions? The answers are both biological and cultural.

Facial expressions are hardwired in the brain

Scientists believe that our brains are  hardwired  to use the facial muscles in specific ways to show our emotions. They suggest that this developed because people needed to live in groups to survive. This neurological phenomenon happens not only in people who can see and imitate the expressions of others but also in people who were born blind.

Facial expressions are both instinctual and learned

Our expressions are instinctual, but we can also learn them from others. Did you ever notice a child's smile that looked identical to a parent's smile? That can happen not only between biological parents and children but also between parents and their adopted children. It's because they tend to imitate their parents’ expressions.

Along with imitating our relatives, we tend to watch others in our culture to learn how to express our emotions. We may meet others in person, watch them on a television show or YouTube video, or see their expression in a photo. When we do, we instinctually understand what they're expressing, and we can learn to express that emotion in the same way.

Which comes first: The expression or the feeling?

We tend to think it's our emotions that determine our facial expressions. However, the facial-feedback hypothesis states that expression can work in the opposite direction. That is, the way we contract our facial muscles may generate emotional feelings within us. The question of whether that happens is still the subject of research studies.

What does the facial-feedback hypothesis mean to me?

How our expressions influence our emotions may pose some interesting questions, but does it have any practical applications? If the facial-feedback hypothesis is true, as research up to the present seems to indicate, there may be several ways to take advantage of the phenomenon. Researchers have found that facial feedback appears to happen during the movement of facial muscles to create expressions, which attenuates ongoing feelings and emotions.

Enjoy life more

Do you ever find yourself in a situation you'd rather avoid? Perhaps you have to be in class or at work when you'd rather be outside enjoying a beautiful day. Maybe you need to interact socially to advance your career or promote your favorite cause, but you'd rather spend the time alone.

If you apply the facial feedback hypothesis in these situations, you might find that you enjoy your time even if you're doing something you'd rather not do. As you smile, happy feelings may follow, allowing you to enjoy these moments wherever you are.

Avoid negative emotions more often

If facial feedback can also cause negative emotions, you may be able to mitigate these feelings or feel them less frequently. If you don't want to feel unhappy, you may try to avoid frowning. If you don’t want to feel angry, you may decide to stop clenching your teeth and decide to modify your expression. If the theory is correct, unpleasant feelings may be far less troublesome.

Have more understanding and control over emotions

It can be healthy and mature to acknowledge your present feelings without wholly giving in to them. You may be able to control distressing emotions, which can improve your mental health. That doesn't mean you never show emotions spontaneously, but you have other options when you need them.

If your emotions sometimes make you feel overwhelmed, facial feedback may help. You can learn valuable techniques from a counselor during online therapy. Aside from teaching you new techniques for controlling your emotions, therapy may help you explore the issues behind those emotions and address any underlying problems.

Online therapy can support you

If you’re thinking about your next steps, online therapy may help you explore your concerns under the guidance of an experienced, licensed counselor. A 2018 study published in the Journal of Anxiety Disorders found that online therapy was equally as effective as traditional in-person counseling; 80% of the trials conducted on computer-delivered therapy sessions saw more than half of the participants showing high rates of satisfaction.

When you choose  online counseling , you can work with a licensed counselor that you choose among thousands of counselors. You can select a therapist who addresses the same types of emotional challenges you're facing, whether you're experiencing anger, sadness, anxiety, or another emotion.

You can also choose a counselor for the type of therapy they offer, whether cognitive behavioral therapy, existential therapy, or dialectical behavior therapy. Their specialties, experience, and educational backgrounds are available for you to read and assess before you set up your first appointment.

Therapist reviews

“Sharon helps you discuss your struggles and then somehow knows the exact words to inspire action. She has helped immensely with my negative self-talk and has brought up my self-esteem a lot.”

“Her guidance throughout this process of change has helped tremendously. When I get off the phone I feel a sense of release of what was clouding my mind. I have tendencies to have negative thoughts and with the techniques she has brought to my attention, I’ve been able to redirect my thoughts to a reality-based point of view. It’s been two weeks and I feel my path with counseling has made an impact already.”

The facial feedback hypothesis is the theory that facial expressions can activate and regulate emotions by influencing the processing of emotional stimuli. By smiling when you’re happy, this hypothesis suggests that you will feel happier. Or, by furrowing your brow when you’re angry, you may feel angrier. This concept was first introduced by Charles Darwin in 1872, but it did not become popular until the 1980s, when the facial feedback hypothesis was defined in the Journal of Personality and Social Psychology . It has since received both criticism and praise as more research is conducted to test this hypothesis. 

What does facial feedback suggest?

Most people believe that we smile when we’re happy, or frown when we’re sad. However, according to the facial feedback hypothesis, the inverse could be true. This would suggest that smiling could cause happiness, and angry facial expressions could cause anger. 

What is the facial feedback hypothesis proposed by James-Lange?

William James and Carl Lange developed the James-Lange Theory of Emotion. According to this theory, physiological changes trigger emotions. For example, if you encountered a rabid dog, your heart rate would rise and you may start perspiring and running away from the dog. James and Lange propose that these physiological changes trigger the emotion and expression of fear, rather than fear triggering the physiological response. Simply put, James and Lange would say you feel afraid because your heart rate has risen, rather than your heart rate rising because you feel afraid. 

The James-Lange Theory has received significant criticism. For example, this theory does not explain why people with limited physiological responses or reduced sensations still experience emotions. 

What is the facial feedback hypothesis of Charles Darwin?

The first of several facial feedback hypotheses was introduced by Charles Dawin in 1872, when he proposed that emotional facial expressions are ubiquitous and innate (not socially learned) in his book, The Expression of the Emotions in Man and Animals. In this book, Darwin observed that emotions intensified when facial muscle regions were engaged , and softened when facial responses were repressed. 

William James and Carl Lange later built on this theory, developing the James-Lange Theory that facial expressions and other physiologic changes generate emotional states.

What is the facial feedback hypothesis replication crisis?

When scientists ask research participants to adopt a voluntary facial action (i.e., by instructing participants to smile or frown), they can unintentionally skew study results. To address this concern, a 1988 study used pens to manipulate facial expression by having participants hold a pen between their teeth or lips, thus inducing smiling or frowning without participant awareness. The researchers then had participants look at a series of cartoons, and found that “smiling” participants reported more positive emotions and found the cartoons more amusing. However, other studies have failed to replicate these results . 

In 17 separate direct replications of the 1988 study , results indicated no significant difference between the “smile” and “frown” groups. 

What are the benefits of facial feedback?

If true, the facial feedback phenomenon suggests that you may feel happier simply by smiling. Therefore, if you’re feeling down, you could boost your mood by reminding yourself to smile. Many people support the notion that “faking it till you make it,” or “turning your frown upside down,” can make you happier. 

However, there is mixed support for the facial feedback hypothesis. Additionally, forcing yourself to disingenuously smile may have a negative impact on your wellbeing. One study found that service workers who felt obligated to smile while interacting with customers experienced heightened rates of excessive alcohol consumption . 

Finally, there have been debates about what nonverbal behaviors, like smiling, actually communicate. While some social psychologists believe that smiling is an expression of happiness, others believe it’s used as a form of social influence to indicate willingness to cooperate with others. 

How has facial feedback effect been supported?

There has been mixed evidence regarding the facial feedback effect. For example, researchers have used the voluntary facial action technique to test the facial feedback hypothesis, instructing participants to induce frowns or smiles in response to positive and negative stimuli, and then rate the pleasantness. They found that facial expressions can reduce the intensity of emotional states , but this effect is generally only present during the actual facial action.

A large 2019 meta-review examined 50 years of research, including 286 studies, found that altering facial expressions has a very small or nonexistent effect on mood . On average, if 100 people smiled, seven may feel happier than they would without smiling .

What best explains the facial feedback effect?

Facial feedback literature was widely popular in the 1980s and 1990s, with particular focus on the vascular theory of facial efference. This theory proposes that facial feedback effects occur when the facial muscles are activated, which may regulate cerebral blood flow and therefore influence emotions . 

How does Botox relate to the facial feedback hypothesis?

Botulinum toxin (Botox) is often injected into the upper region of the face, where it can reduce dynamic creases utilized in some expressive behaviors, such as anger and shock. Some doctors believe that Botox injections could limit the ability to frown , and thus may reduce negative facial feedback effects, leading to more positive emotional states overall. 

What is the facial feedback hypothesis on Quizlet?

By creating a free Quizlet account, you can review user-generated study questions and flashcards that cover the facial feedback hypothesis and other psychology concepts . 

• Exploring the different types of humor and their effects on personalities Medically reviewed by Majesty Purvis , LCMHC
• What are alpha brain waves? Medically reviewed by Arianna Williams , LPC, CCTP
• Relationships and Relations

Facial Feedback Hypothesis: The Power of Facial Expressions in Shaping Emotions

The facial feedback hypothesis is a fascinating concept in psychology that suggests our facial expressions can influence our emotions. According to this theory, when we smile, it not only reflects happiness but also has the power to actually make us feel happier. Similarly, frowning might intensify feelings of sadness or anger.

This hypothesis proposes that our facial muscles send signals to our brain, which then interprets those signals and triggers corresponding emotional responses. By altering our facial expressions, we may be able to manipulate and regulate our emotions to some extent.

Research into the facial feedback hypothesis has yielded intriguing results. For example, studies have shown that participants who held a pen between their teeth (which activates the smiling muscles) rated cartoons as funnier compared to those who held a pen between their lips (which inhibits smiling). These findings suggest a strong connection between facial expressions and emotional experience.

Understanding the impact of facial feedback on emotions can have important implications for various fields such as therapy, communication, and even daily interactions. By recognizing how our own expressions affect our mood and considering the influence of others’ expressions on their emotions, we can enhance empathy and improve interpersonal relationships.

In conclusion, the facial feedback hypothesis proposes that there is a bidirectional relationship between our facial expressions and emotions. The way we use or modify our faces can potentially shape how we feel internally. As I delve deeper into this topic, let’s explore the research behind this intriguing theory and its practical applications in more detail.

What is the Facial Feedback Hypothesis?

The Facial Feedback Hypothesis suggests that our facial expressions can influence and even shape our emotions. According to this hypothesis, when we make certain facial expressions, such as smiling or frowning, it can trigger corresponding emotional responses within us. In other words, our facial muscles send signals to our brain that then impact how we feel.

For example, have you ever noticed that when you force yourself to smile, even if you’re not feeling particularly happy at the moment, you start to feel a bit better? This phenomenon aligns with the Facial Feedback Hypothesis. By activating the muscles associated with smiling, your brain receives feedback that there may be something positive happening and responds by releasing feel-good chemicals like endorphins.

On the other hand, furrowing your brow or scowling might lead to feelings of anger or frustration. The tension in these facial muscles sends signals to your brain that something negative is occurring, which can intensify those negative emotions.

Research has shown various interesting findings related to the Facial Feedback Hypothesis. One study conducted by psychologist Fritz Strack and colleagues in 1988 involved participants holding a pen either between their teeth (forcing a smile-like expression) or between their lips (preventing any significant facial movement). The participants who held the pen between their teeth reported finding cartoons funnier compared to those who held it between their lips. This experiment provided evidence supporting the idea that facial expressions can impact emotional experiences.

Another study published in 2012 by Joshua Davis and Ann Senghas explored how altering participants’ facial expressions affected their emotional reactions while watching emotionally charged videos. Participants were instructed to either hold chopsticks with their teeth (creating a smile-like expression) or hold them with their lips (creating a frown-like expression). Interestingly, those who held the chopsticks with their teeth reported feeling more positive emotions than those who held them with their lips.

These studies and others like them demonstrate the potential power of facial expressions in shaping our emotional experiences. While there is still much to learn about the intricacies of this hypothesis, it offers intriguing insights into the connection between our facial muscles and our emotions. It highlights the idea that our expressions not only reflect how we feel but also have the ability to influence how we feel.

The Role of Facial Expressions in Emotion

When it comes to understanding human emotions, facial expressions play a crucial role. Our faces are not only a canvas for displaying our feelings but also an essential means of communication. The facial feedback hypothesis suggests that the expressions we make with our faces can actually influence and even intensify our emotional experiences.

Think about it – when you’re feeling joyful, your face naturally lights up with a smile. But what if I told you that the reverse is also true? That by intentionally putting on a smile, you can actually boost your mood and experience genuine happiness? It may sound surprising, but research has shown that simply forcing yourself to smile can lead to increased positive emotions.

A classic study conducted by psychologist Paul Ekman found that participants who held a pen between their teeth (forcing them into a smiling expression) while watching cartoons reported greater amusement compared to those who held the pen with their lips (preventing them from smiling). This demonstrates how our facial expressions can influence the way we perceive and interpret emotions.

Furthermore, studies have indicated that specific facial expressions may be universally recognized across different cultures. For example, furrowed brows and tight lips typically convey anger or frustration, while raised eyebrows and widened eyes often indicate surprise or fear. These universal expressions suggest an innate connection between facial movements and emotional states.

In addition to influencing our own emotions, facial expressions also play a vital role in social interactions. They serve as nonverbal cues that help us understand the emotional states of others and enable empathy and effective communication. Imagine trying to navigate through life without being able to read subtle changes in someone’s face – it would certainly make understanding others’ feelings much more challenging.

In conclusion, the role of facial expressions in emotion cannot be underestimated. Our faces act as powerful tools for both expressing and experiencing emotions. By recognizing the impact of our own facial expressions on our emotional well-being, we can harness this knowledge to shape our moods and enhance our interpersonal connections. So, the next time you find yourself in need of a mood boost, try putting on a smile – it might just make all the difference.

The Connection between Facial Feedback and Mood

Let’s dive into the fascinating connection between facial feedback and mood. It’s truly intriguing how our facial expressions can influence our emotional state. As we explore this topic further, you’ll discover the impact that simple facial movements can have on shaping our moods.

Have you ever noticed that when you smile, even if it’s forced, you start to feel a little bit happier? That’s because our facial muscles send signals to our brain, triggering the release of neurotransmitters associated with positive emotions. So, by simply putting on a smile, even when we’re not genuinely happy, we can actually boost our mood.

On the flip side, frowning or expressing negative emotions through our facial expressions can have the opposite effect. When we furrow our brows or tighten our jaw in frustration or sadness, it sends signals to the brain that reinforce those negative feelings. This is why sometimes we find ourselves feeling even more down when we allow ourselves to dwell on negative thoughts.

To illustrate this connection between facial feedback and mood, let me share an interesting study conducted at a university campus. Researchers asked participants to hold a pencil horizontally in their mouths without letting their lips touch it – essentially simulating a smile – while watching funny videos. Surprisingly, those who held the pencil in this way reported finding the videos funnier compared to those who held it vertically – mimicking a frown.

Another example comes from everyday life experiences. Think about how often people use phrases like “putting on a brave face” or “grin and bear it”. These idioms reflect an intuitive understanding of how altering one’s facial expression can impact their emotional state.

In conclusion, there is solid evidence supporting the link between facial feedback and mood. By consciously manipulating our facial expressions towards positivity (even if initially artificial), we can influence and uplift our overall emotional well-being. So next time you’re feeling down, try putting on a smile and see how it affects your mood.

Empirical Evidence Supporting the Facial Feedback Hypothesis

Let’s delve into the empirical evidence that supports the Facial Feedback Hypothesis. Numerous studies have been conducted to explore the relationship between facial expressions and emotions, providing compelling support for this intriguing theory. Here are a few examples:

• The Pen-in-Mouth Experiment: In one classic study, participants were instructed to hold a pen either between their teeth (activating smiling muscles) or between their lips (preventing smiling muscles from engaging). The results showed that those who held the pen in their teeth reported feeling happier than those with the pen in their lips. This suggests that even mimicking a smile can influence our emotional state.
• Botox and Emotional Experience: Another line of research involves studying individuals who have received Botox injections, which temporarily paralyze certain facial muscles and limit their ability to express emotions fully. Studies have found that these individuals report experiencing less intense negative emotions compared to those without Botox treatments, supporting the idea that facial expressions play a role in how we experience emotions.
• Mirror Neurons and Emotional Contagion: Mirror neurons are brain cells that fire both when we perform an action and when we observe someone else performing the same action. They contribute to our ability to empathize with others’ emotions by mirroring their facial expressions internally. Research has shown that mirroring others’ expressions can activate corresponding emotional states within ourselves, further strengthening the link between facial feedback and emotional experience.
• Cultural Universality of Facial Expressions: Cross-cultural studies examining facial expressions of emotion have consistently revealed remarkable similarities across different cultures worldwide. This suggests that certain universal patterns of facial expression are innate rather than learned behaviors, reinforcing the notion that our faces play a fundamental role in communicating and experiencing emotions.
• Neuroimaging Studies: Advanced brain imaging techniques such as functional magnetic resonance imaging (fMRI) have provided valuable insights into how our brains process emotions. These studies have shown that when we observe facial expressions, specific regions of the brain associated with emotional processing and empathy are activated, further supporting the connection between facial feedback and emotion regulation.

The empirical evidence in support of the Facial Feedback Hypothesis is both diverse and compelling. From experimental manipulations to neuroimaging studies, these findings consistently suggest that our facial expressions not only reflect our emotions but also influence how we feel. This research opens up exciting possibilities for understanding and harnessing the power of our faces to enhance emotional well-being.

Critiques and Controversies Surrounding the Hypothesis

Now let’s delve into some of the critiques and controversies surrounding the facial feedback hypothesis. While this theory has gained significant attention and support, it is not without its fair share of skepticism and debate. Let’s explore a few key examples:

• Lack of Consistency in Research Findings: One major criticism revolves around the inconsistency in research findings related to the facial feedback hypothesis. Some studies have provided evidence supporting the idea that facial expressions can influence emotions, while others have failed to replicate these results. This discrepancy has led some researchers to question the reliability and validity of the hypothesis.
• Methodological Limitations: Another point of contention lies in the methodological limitations associated with studying facial expressions and their impact on emotions. Conducting experiments in this field can be challenging due to factors such as subjective interpretation, participant biases, and difficulty controlling all variables involved. These limitations raise concerns about whether existing research truly captures a comprehensive understanding of how facial expressions influence emotional experiences.
• Alternative Explanations: Critics also propose alternative explanations for why certain studies have shown effects consistent with the facial feedback hypothesis. For instance, they argue that changes in emotional states observed during experiments could be attributed to other factors unrelated to specific muscle movements or facial expressions. These alternative explanations highlight the need for further investigation before drawing definitive conclusions about how much influence our facial muscles truly have on our emotions.
• Cultural Differences: Another aspect worth considering is cultural variation in interpreting and expressing emotions through facial expressions. What may be perceived as a universal response in one culture might differ significantly across different societies or regions. This cultural variability raises questions about whether findings from studies conducted within specific cultural contexts can be generalized to a global population.
• Publication Bias: Finally, critics draw attention to publication bias within academic journals, which tends to favor positive results over null findings or contradictory evidence regarding theories like the facial feedback hypothesis. This bias can create an inflated perception of support for the hypothesis and hinder a more balanced evaluation of its validity.

It’s important to acknowledge these critiques and controversies surrounding the facial feedback hypothesis as they shed light on potential limitations and areas for further exploration. By critically examining the existing research, we can continue to refine our understanding of how facial expressions may or may not influence our emotional experiences.

Implications for Emotional Regulation and Well-being

When it comes to emotional regulation and overall well-being, the facial feedback hypothesis offers intriguing implications. By understanding how our facial expressions can influence our emotions, we gain valuable insights into managing and improving our mental state.

Enhancing Positive Emotions:

Research suggests that deliberately adopting positive facial expressions can actually enhance our experience of positive emotions. For instance, by smiling even when we don’t feel particularly happy, we may trick our brain into thinking that we are indeed happy. This simple act of “putting on a smile” can have a profound impact on boosting mood and promoting feelings of joy and contentment.

Managing Negative Emotions:

On the flip side, the facial feedback hypothesis also sheds light on how regulating our facial expressions can help manage negative emotions. When faced with distressing situations or stressors, consciously relaxing the muscles in our face or adopting a neutral expression can help reduce intensities of anger, sadness, or anxiety. By controlling our facial expressions, we may be able to exert some control over the intensity and duration of negative emotional states.

Improving Self-awareness:

Paying attention to our own facial expressions can serve as a powerful tool for self-reflection and self-awareness. By becoming more attuned to how certain expressions correspond to specific emotions within ourselves, we become better equipped at recognizing and understanding our own emotional states. This heightened awareness allows us to respond more effectively to challenging situations by employing strategies tailored to regulate those emotions.

Strengthening Social Connections:

Our facial expressions play a vital role in social interactions as well. The ability to accurately interpret others’ emotions based on their facial cues is crucial for effective communication and building strong relationships. Understanding the link between facial expressions and emotions enables us not only to decipher others’ feelings but also empowers us with the knowledge of how our own expressions may impact those around us positively or negatively.

In conclusion, the implications of the facial feedback hypothesis for emotional regulation and well-being are vast. By recognizing the power of our facial expressions and leveraging them consciously, we can enhance positive emotions, manage negative emotions, improve self-awareness, and strengthen social connections. Incorporating this knowledge into our daily lives may lead to greater emotional balance and overall well-being.

Applications in Psychology and Therapy

One of the fascinating aspects of the facial feedback hypothesis is its potential applications in psychology and therapy. This theory suggests that our facial expressions not only reflect our emotions but also have the power to influence them. By understanding and harnessing this connection, professionals in the field can explore various ways to improve mental well-being and enhance therapeutic interventions.

Here are a few examples showcasing how the facial feedback hypothesis has been applied in psychology and therapy:

• Emotion Regulation: The ability to regulate emotions is crucial for psychological well-being. Research has shown that intentionally altering one’s facial expressions can impact emotional experiences. For instance, adopting a smiling expression, even when feeling down, may help elevate mood levels. Therapists can incorporate techniques such as “smile therapy” or encouraging clients to engage in activities that naturally elicit positive expressions to aid in emotion regulation.
• Cognitive Behavioral Therapy (CBT): CBT focuses on challenging negative thought patterns and replacing them with more positive ones. Facial feedback techniques can be integrated into CBT sessions by encouraging clients to use their facial muscles deliberately to express positive emotions during challenging situations. This practice aims to reinforce adaptive cognitive processes by aligning facial cues with desired emotional states.
• Nonverbal Communication: Facial expressions play a vital role in nonverbal communication, conveying information about one’s thoughts and feelings without words. Therapists can utilize this aspect by closely observing clients’ facial expressions during sessions, looking for signs of discomfort, tension, or hidden emotions that may provide valuable insights into their psychological state.
• Body-Mind Connection: The mind-body connection is a fundamental concept within psychology and therapy. The facial feedback hypothesis strengthens this relationship by highlighting how changes in our physical body (specifically through facial muscles) can impact our mental state. By incorporating exercises like relaxation techniques involving specific facial muscle groups, therapists aim to promote overall relaxation and reduce stress levels.
• Social Anxiety Treatment: Social anxiety disorder is characterized by intense fear of social situations. Studies have shown that individuals with social anxiety often exhibit facial expressions associated with negative emotions, such as fear or disgust, even in neutral or positive contexts. Therapists can employ techniques to help clients consciously modify their facial expressions, encouraging more relaxed and open postures to reduce anxiety and improve social interactions.

These examples illustrate the various ways in which the facial feedback hypothesis has found practical applications within the field of psychology and therapy. By exploring the influence of facial expressions on emotional experiences, professionals can develop innovative interventions to support individuals in improving their mental well-being and fostering healthier emotional responses.

Based on the research and evidence presented in this article, it is clear that the facial feedback hypothesis offers valuable insights into the connection between facial expressions and emotional experiences. By examining how our facial muscles influence our emotions, researchers have shed light on the complex interplay between our bodies and minds.

Here are a few key examples that highlight the significance of the facial feedback hypothesis:

• Facial Expressions and Emotional States: The studies discussed throughout this article consistently demonstrate a strong correlation between specific facial expressions and corresponding emotional states. For instance, when individuals were asked to hold a pen with their teeth (mimicking a smile), they reported feeling happier compared to those who held it with their lips (mimicking a frown). These findings provide compelling evidence for the impact of facial muscle activity on emotional experiences.
• Mirror Neurons and Empathy: Research has shown that observing someone else’s facial expression can activate mirror neurons in our brains, leading to an empathetic response. This suggests that not only do our own emotions influence our facial expressions, but also that witnessing others’ expressions can elicit similar emotional responses within ourselves. Such findings support the notion that facial feedback plays a crucial role in social interactions and empathy.
• Therapeutic Applications: The understanding of how facial expressions affect emotions has practical implications in various therapeutic settings. Techniques such as cognitive reappraisal, where individuals consciously alter their interpretations of situations to regulate emotions, often involve modifying one’s facial expression as well. Additionally, therapies like laughter yoga utilize forced laughter as an intervention to induce positive emotions through changes in facial muscle activity.

In conclusion, the research surrounding the facial feedback hypothesis provides compelling evidence for the bidirectional relationship between our faces and emotions. Understanding this connection can have far-reaching implications for fields such as psychology, neuroscience, and even therapy techniques aimed at enhancing well-being. As we continue to explore this fascinating area of study, further insights into how our facial expressions shape our emotional experiences are likely to emerge.

Related Posts

Can You Be in Love with Two People: Unpacking the Complexity of Human Emotions

Emotions vs Feelings: Understanding the Intricate Differences

Nonverbal behavior and the theory of emotion: the facial feedback hypothesis

• PMID: 7381683
• DOI: 10.1037//0022-3514.38.5.811

The facial feedback hypothesis, that skeletal muscle feedback from facial expressions plays a causal role in regulating emotional experience and behavior, is an important part of several contemporary theories of emotion. A review of relevant research indicates that studies reporting support for this hypothesis have, without exception, used within-subjects designs and that therefore only a restricted version of the hypothesis has been tested. Also, the results of some of these studies must be questioned due to demand characteristics and other problems. It is suggested that visceral feedback may make a more direct contribution to emotional processes than facial feedback does and that the "readout" functions of facial expressions are more important than any feedback functions.

• Emotions / physiology*
• Facial Expression*
• Facial Muscles / physiology*
• Nonverbal Communication

Cannon-Bard Theory of Emotion: Definition and Examples

Charlotte Nickerson

Research Assistant at Harvard University

Undergraduate at Harvard University

Charlotte Nickerson is a student at Harvard University obsessed with the intersection of mental health, productivity, and design.

Learn about our Editorial Process

Saul Mcleod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul Mcleod, PhD., is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Olivia Guy-Evans, MSc

Associate Editor for Simply Psychology

BSc (Hons) Psychology, MSc Psychology of Education

Olivia Guy-Evans is a writer and associate editor for Simply Psychology. She has previously worked in healthcare and educational sectors.

On This Page:

Key Takeaways

• The primary argument of the Cannon-Bard theory of emotion is that emotions trigger affective “feelings” and physiological responses to stimuli simultaneously in different regions of the brain. This stands in contrast to the James-Lange theory of emotion, which posits that people and animals feel emotions because they consciously process their physiological responses to stimuli.
• The Cannon-Bard theory of emotion differentiates between feelings associated with the sympathetic nervous system (fight or flight responses) and the parasympathetic nervous system (calm responses), and Cannon believed that sympathetic and parasympathetic responses could not happen simultaneously.
• The Cannon-Bard and James Lange theories of emotion have greatly influenced modern research into emotional processing and the brain; however, both theories have garnered great criticism for their overgeneralization of emotion and contradictions between theory and evidence dating as far back as Cannon’s own research.

According to Cannon-Bard theory of emotion, physiological arousal and emotional experience occur simultaneously, yet independently. This theory was proposed in the 1920s and early 1930s by Walter B. Cannon and Philip Bard.

In short, the Cannon-Bard theory of emotion, also known as the Thalamic theory of emotion, states that the lower part of the brain, what neurologists call the thalamus , controls emotional experience.

Meanwhile, the higher part of the brain, the cortex , controls emotional expression. These feelings (through the thalamus) and physical reactions (through the cortex) occur at the same time. For example, seeing a spider may trigger both a physical response (such as jerking back one’s hand) and an affective, emotional one (such as a feeling of fear).

The Cannon-Bard Theory of Emotion represented a shift from the James-Lange Theory of emotion to one which studied central brain mechanisms as the cause of emotions; however, work as far back as the 1860s posited physiological causes for emotion (Dror, 2014).

How Does the Cannon-Bard Theory Work?

The main idea of Cannon’s approach to emotions is that people react to emotional stimuli but that two separate parts of the brain control the conscious feeling of emotion and the body’s physiological response.

Cannon reviewed research on emotions in both animals and people with brain damage, as well as conducted his own experiments. He concluded that the thalamus was the brain region most involved in physiological emotion, while the cortex is responsible for controlling and inhibiting it.

This is supported by modern neuroscience, which believes that the thalamus is responsible for relaying sensory and motor signals to the cerebral cortex, which handles processes related to thought, consciousness, reasoning, and memory.

Stressful Exam

Say that a student is about to take a high-stakes midterm exam that he has not studied for. This stressful stimulus (an upcoming exam) triggers two brain regions separately: the cortex and the thalamus.

The cortex triggers a physiological response to the emotion. For example, the student may be sweating, his heart may race, and his fingers may tremble as he picks up his pencil.

According to the Cannon-Bard theory of emotion, these are physiological responses to stress triggered by the sympathetic nervous system, and the triggering of the sympathetic nervous system means that the parasympathetic nervous system cannot be triggered.

Meanwhile, the stimulus triggers the student’s thalamus separately, leading to the conscious, affective feeling of stress.

The student is not stressed because he is trembling, according to Cannon-Bard, but because his thalamus has been activated by an impending exam.

This emotional reaction would be separate and independent of the physiological arousal, even though they co-occur.

Purring Cat

Taking an example from Bard’s 1936 letter to Cannon, consider a resting cat who is being petted by its owner. As the owner pets the cat, the cat may purr and relax its muscles or slowly fall asleep.

This purring is a physiological response of the parasympathetic nervous system to the stimulus of being a pet, and, according to the Cannon-Bard theory of emotion, exists entirely separately from the cat perceiving a conscious emotion (such as calmness or comfort).

According to the Cannon-Bard theory of emotion, the cat is not calm because it is purring and has relaxed muscles; rather, it is calm because the stimulus of petting has activated its thalamus, allowing the cat to have the conscious, affective emotion of calm.

Comparison to Other Theories

James-Lange theory of emotion came to influence a century of empirical emotional research and rebuttals of the James-Lange theory of emotion, notably Cannon-Bard’s 1927 critique, have spurned long-standing debates in neuroscience and psychology (Lang, 1994).

James-Lange Theory of Emotion

Both the James-Lange and Cannon-Bard Theories of emotion investigated whether emotions originated from a source central or peripheral to the nervous system (Meiselman, 2016).

The James-Lange theory of emotion is ultimately a peripheral approach to understanding emotion. This means that rather than first perceiving some stimulus that could elicit an emotion, experiencing an emotion, and then experiencing some bodily reaction in response, James-Lange assumes that the bodily reactions themselves elicit conscious emotions.

Cannon offers five objections to the James-Lange theory of emotion:

1.) The latency of physiological responses to emotion is too long to account for the immediacy of emotional behavior;

2.) Artificially inducing bodily reactions to emotion does not in itself produce emotions;

3.) The viscera are so-called “insensitive structures;”

4.) Visceral changes are the same between emotions and

5.) Interrupting the feedback that occurs because of visceral bodily emotions does not influence emotional behavior (Fehr and Stern, 1970).

To illustrate the contrast from which these five objections arise, consider somebody who is afraid of dogs. The dog barks within range of the person, and the person is aroused, according to the James-Lange theory of emotion, autonomically.

This person could have a physical response such as trembling. Because of that autonomic arousal (the trembling), the person has a conscious feeling of fear. The person feels afraid because they tremble.

Meanwhile, according to the Cannon-Bard theory of emotion, the dog’s bark triggers two separate regions of the brain: the thalamus and the cortex.

Because the dog triggers the cortex, which controls physical reactions, the person trembles. However, the person feels fear because the bark triggers the thalamus, which controls the conscious feelings of emotion.

In theory, if that person’s thalamus or cortex were dysfunctional, the person may be able to feel fear without trembling or tremble without feeling fear. The dog makes the person tremble and feel afraid, but these are reactions stemming from two entirely different parts of the brain.

Considering his objections, Canon introduced a specificity model of emotions (Dror, 2014). He distinguished different classes of emotions by whether they affected the sympathetic or parasympathetic nervous system.

Emotions that affect the sympathetic nervous system are associated with fight or flight responses , and those that activate the parasympathetic nervous system restore the body to a state of calm (Waxenbaum, 2021). The anatomical, physiological, and metabolic differences between the emotions expressed by the sympathetic and parasympathetic nervous systems create a difference in how people consciously label these emotions (Cannon, 1914).

In contrast to the Cannon-Bard and James-Lange theories of emotion, the so-called “common sense” theory of emotion posits that someone has an emotional response, and this response triggers a physiological reaction.

For example, someone can see a barking dog and consciously feel fear. As a result, they tremble (Meiselman, 2016).

Singer’s Two-Factor Theory of Emotion

In the Schachter-Singer Two Factor Theory of Emotion , one does a conscious cognitive appraisal of their physiological response, labels that response, and feels the emotion that results. Neither physiological nor cognitive arousal in itself is enough to elicit an emotion, according to Schacter-Singer.

The Cannon-Bard theory maintains that emotional experience occurs simultaneous to and independent of physiological arousal.

The Schachter-Singer two-factor theory suggests that physiological arousal receives cognitive labels as a function of the relevant context and that these two factors together result in an emotional experience.

To repeat the example of a barking dog, a person may tremble or feel an increased heart rate. After detecting that they are trembling, the person can then label the physiological response (“I must be scared!”) and only then consciously feel the emotion.

A person, according to the Two Factor Theory of Emotion, must appraise their physiological responses in order to experience emotion (Meiselman, 2016).

Zajonc-LeDoux Theory of Emotion

The Zajonc-LeDoux theory of emotion says that emotional reactions exist separately from cognitive labels on emotional situations.

According to this theory, some emotions that have evolutionarily been necessary (e.g., anger or fear) are activated through quicker pathways than others.

Certain emotions can happen instantly without an active cognitive appraisal. For example, someone can be startled by the bark of a dog before labeling it as a threat (Meiselman, 2016).

Key Research

In 1925, Cannon and Britton introduced a way of studying emotions through a cat whose cortex had been removed from his brain (Cannon and Britton, 1925).

From this cat, Cannon and Britain described a new type of emotion — “sham rage.”

At the time, Cannon noticed that laboratory animals did not consistently develop the desired emotional reactions to stimuli consistently and dependently, so Cannon removed the region of the brain — the cortex — thought to inhibit emotions, allowing Cannon to study a stream of emotions over longer periods of time.

This decorticated cat came to underpin Cannon’s physiological model of emotional experience.

In the mid-1920s, Philip Bard was a doctoral student studying under Walter Bradford, a physiologist at Harvard University, and two years after introducing this method for studying the physiology of emotions, Cannon suggested that Philip Bard section the cat’s brain in order to discover which parts of the brain were responsible for emotional expression (Bard, 1973).

From this, Cannon-Bard discovered that the thalamus generated affective emotions.

Many neurologists have initiated empirical research in order to determine how and in which order emotion affects various regions of the body.

The first of these researchers was Allport, who devised the facial feedback hypothesis (Laird, 1984). In short, the facial feedback hypothesis postulated that one’s facial expressions directly affect one’s emotional experience. For example, someone who tries to smile will feel happier because they have a smiling expression.

To Tomkins (1962-1963), emotions are “sets of muscles and glandular responses located in the face.”

Later, researchers suggested embodied theories of emotion. According to embodied theories of emotion, simply contracting the facial muscles associated with a particular emotion can intensify or elicit the emotion associated with those contractions, even when participants are not contracting those muscles consciously (Niedenthal, 2007; Soussignan, 2002).

Cannon (1927) conducted several animal tests to disprove James-Lange’s peripheralist approach to emotion (Meiselman, 2016). For example, Cannon conducted tests where he artificially induced certain bodily reactions and observed that these in themselves did not elicit emotions (Meiselman, 2016).

Researchers such as Fehr and Stern (1970), among many others, have criticized these tests. One of the predominant criticisms of the Cannon-Bard Theory of Emotion is that the theory assumes that physical reactions do not influence emotions.

For example, Cannon-Bard would assume that someone would not necessarily feel happy if the facial muscles creating a smile were triggered.

However, a large body of research on facial expressions and emotions suggests that this assumption is not true.

Some scholars, such as Dror (2014), have argued that Cannon’s own studies have contradicted his theory of emotion.

For example, one large component of Cannon’s physiological theory of emotion was the belief that either the sympathetic or parasympathetic nervous system could be activated by emotion, but the activation of these systems was mutually exclusive.

Relatively common phenomena — such as the contraction of the bladder and rectum during intense emotional stress — would confuse Cannon and his followers because these represented the simultaneous activation of the sympathetic and parasympathetic nervous systems (Cannon, 1914).

Scholars, and indeed Cannon himself, were aware of contradictions and flaws in his studies of the decorated cat as a model for physiological and emotional experience (Dror, 2014).

For example, as Dror (2014) emphasizes, Cannon and Britton included clear evidence of parasympathetic activation, such as contraction of the rectum and occasionally defecation, in their descriptions of the cat experiencing “sham rage” – a supposedly sympathetic, fight or flight emotion (Cannon and Britton, 1925).

Many other physiologists in the 1930s would confirm this evidence that “shame rage” activated both the sympathetic and parasympathetic nervous systems (Beattie, 1932; Beebe-Center and Stevens, 1938).

The second major flaw of Cannon’s theory, according to Dror’s 2014 review, was that there was a lack of anatomical proof for two major aspects of the Cannon-Bard theory of emotion.

The stipulation of the Cannon-Bard model that the optic thalamus was the region that organized emotional expressions, as well as the belief that the thalamus itself was the source of the affective experience of emotion, lacked anatomical proof (Cannon, 1927).

Thirdly and lastly, scholars have criticized Cannon’s theory for its overextension beyond the evidence provided by the Cannon-Bard experiments.

Although Cannon-Bard studied the decoriated cat’s “sham rage” exclusively, Cannon would go on to extend his model of emotions to joy, grief, and disgust long before Philip Baard himself would send a letter to Cannon observing pleasure in decorticated cats (such as pleasure when being petted) (Dror, 2014).

Bard, P. (1973). The ontogenesis of one physiologist. Annual Review of Physiology, 35 (1), 1-16.

Beattie, J. (1932). Hypothalamic mechanisms. Canadian Medical Association Journal, 26(4), 400.

Beebe-Center, J., & Stevens, S. (1938). The emotional responses: changes of heart-rate in a gun-shy dog. Journal of Experimental Psychology, 23 (3), 239.

Bernard, C. (1866). Leçons sur les propriétés des tissus vivants: Germer Baillière.

Cannon, W. B. (1914). The interrelations of emotions as suggested by recent physiological researches. The American Journal of Psychology, 25 (2), 256-282.

Cannon, W. B. (1927). The James-Lange theory of emotions: A critical examination and an alternative theory. The American Journal of Psychology, 39 (1/4), 106-124.

Cannon, W. B., & Britton, S. W. (1925). Studies on the conditions of activity in endocrine glands: XV. Pseudaffective medulliadrenal secretion. American Journal of Physiology-Legacy Content, 72 (2), 283-294.

Coppin, G., & Sander, D. (2021). Chapter 1 – Theoretical approaches to emotion and its measurement. In H. L. Meiselman (Ed.), Emotion Measurement (Second Edition) (pp. 3-37): Woodhead Publishing.

Dalgleish, T., Dunn, B. D., & Mobbs, D. (2009). Affective neuroscience: Past, present, and future. Emotion Review, 1 (4), 355-368.

Darwin, C. (1872). The expression of the emotions in man and animals by Charles Darwin: John Murray.

de Cyon, E. (1873). Principes d”électrothérapie: Baillière.

Dror, O. E. (2013). The Cannon–Bard Thalamic Theory of Emotions: A Brief Genealogy and Reappraisal. Emotion Review, 6 (1), 13-20. doi:10.1177/1754073913494898

Durant, J. R. (1981). The beast in man: An historical perspective on the biology of human aggression. The biology of aggression, 17-46.

Ellsworth, P. C. (1994). William James and emotion: is a century of fame worth a century of misunderstanding? Psychological Review, 101 (2), 222.

Fehr, F. S., & Stern, J. A. (1970). Peripheral physiological variables and emotion: the James-Lange theory revisited. Psychological Bulletin, 74 (6), 411.

Friedman, B. H. (2010). Feelings and the body: The Jamesian perspective on autonomic specificity of emotion. Biological Psychology, 84 (3), 383-393.

Laird, J. D. (1984). The real role of facial response in the experience of emotion: a reply to Tourangeau and Ellsworth, and others.

Lanska, D. J. (2014). Cannon, Walter Bradford. In M. J. Aminoff & R. B. Daroff (Eds.), Encyclopedia of the Neurological Sciences (Second Edition) (pp. 580-583). Oxford: Academic Press.

Meiselman, H. L. (2016). Emotion measurement: Woodhead publishing.

Niedenthal, P. M. (2007). Embodying emotion. Science, 316 (5827), 1002-1005.

Sherrington, C. S. (1900). Experiments on the value of vascular and visceral factors for the genesis of emotion. Proceedings of the Royal Society of London, 66 (424-433), 390-403.

Soussignan, R. (2002). Duchenne smile, emotional experience, and autonomic reactivity: a test of the facial feedback hypothesis. Emotion, 2 (1), 52.

Tomkins, S. (1963). Affect imagery consciousness: Volume II: The negative affects: Springer publishing company.

Waxenbaum, J. A., Reddy, V., & Varacallo, M. (2019). Anatomy, autonomic nervous system.

Further Information

• Cannon, W. B. (1927). The James-Lange theory of emotions: A critical examination and an alternative theory. The American journal of psychology, 39(1/4), 106-124.
• Cannon, W. B. (1914). The interrelations of emotions as suggested by recent physiological researches. The American Journal of Psychology, 25(2), 256-282.
• Barrett, L. F. (2012). Emotions are real. Emotion, 12(3), 413.
• Ellsworth, P. C. (1994). William James and emotion: is a century of fame worth a century of misunderstanding?. Psychological review, 101(2), 222.

Your Facial Expressions Can Impact Your Mood

Facial Feedback Research

The following studies are samples from the Facial Feedback literature review by Katherine Wright and published in 2022 in the Undergraduate Research Journal of the University of Utah. To learn more about the following studies (including statistics) and additional Facial Feedback research, read the full Facial Feedback literature review .

Autonomic Nervous System Activity Distinguishes Emotions

Paul Ekman and Wallace Friesen (1978) developed the Facial Activation Coding System (FACS) in the 1970s to assist researchers in identifying the emotional expression of study subjects. They identified 46 action units that comprise common facial expressions and created a manual so others can identify the emotions of subjects in photographs, films, and live research. Ekman and Friesen both experienced strong physical sensations while contracting their facial muscles into facial expressions during the development of the FACS. This experience led them to develop the theory that “voluntary production of emotional facial configurations would produce emotion-specific patterns of autonomic activity” (Levenson et al. 1990).

A few years later, Ekman, Levenson, and Friesen (1983) studied the autonomic responses of six universal emotions: surprise, disgust, sadness, anger, fear, and happiness. The researchers hypothesized that they would find differences in the autonomic responses between the six emotions. Sixteen subjects contracted specific muscles resulting in a prototypical emotional expression and sensations of the emotion. The subjects also “relived” a personal experience relating to a specific emotion. Heart rate, skin temperature, sweating, and forearm muscle tension of the subjects were measured during these tasks.

Ekman, Levenson, and Friesen found a significant increase in heart rate for anger, fear, and sad emotional experiences , but only anger showed a significant difference in skin temperature. Surprisingly, the autonomic responses created by facial expressions were stronger than those created by relived experiences , and support the sufficiency hypothesis that facial expressions, by themselves, can instigate an emotional experience.

Inhibited or Facilitated Smiles

Strack, Martin, and Stepper’s (1988) facial feedback research also supported the sufficiency hypothesis. Ninety-two study subjects were told that injured or disabled people may need to hold a pen in a non-standard way, so the study subjects held a pen in three different ways as they performed simple tasks, including rating the humor of cartoons:

• with lips only (inhibiting a smile)
• with teeth only (facilitating smile)
• with their non-dominant hand—(Figure 2).

The researchers discovered that subjects that viewed the cartoons with the teeth only (which facilitated a smile) rated the cartoons significantly funnier than those who held the pen with lips only (which inhibited a smile) or with the non-dominant hand. This FFH research was one of the first to successfully blind the subjects to the actual purpose of the study, which increased confidence in the results.

Eyebrow Placement Implicates Facial Feedback Correlation with Emotion

Michael Lewis (2012) further explored the facial feedback hypothesis in three experiments. In his first experiment, 54 participants were told the research regarded concentration methods as a cover story. The participants were divided into three groups. One group was instructed to raise their eyebrows high, another group to lower their eyebrows, and a third group was told to keep their eyebrows in the middle. All groups took the Irritability-Depression-Anxiety (IDA) questionnaire. Researchers discovered that the eyebrows-down group scored significantly higher on the IDA for depression and anxiety but not for irritability. Lewis suggested that these results support the idea that BTX treatments could reduce depression by suppressing frowning capability.

The cover story from experiment one was used in a second experiment. The facial feedback influence of surprise or disbelief was explored as 24 subjects either raised their eyebrows as high as possible, lowered their eyebrows as low as possible, or held them constant while rating the plausibility of ten statements that they were told were “facts.” Eyebrow placement proved to impact the emotional impact of the experience. Those with raised eyebrows were more surprised by the “facts.”

The third experiment examined whether facial expressions of disgust would impact the evaluation of odors. Thirty-three subjects were told that researchers were examining how facial muscle activity moved smells to different parts of the nasal cavity. Subjects then forced their noses as high as possible, as low as possible (by lowering the lower lip), or by keeping the nose relaxed as they smelled boxes of mostly unpleasant odors, such as “urine,” “farmyard,” and “vomit.” Once again, facial manipulation of the nose impacted the pleasantness ratings of the smells. Subjects that forced their noses high, creating an expression of disgust, rated the odors as more unpleasant than those who held their noses low or relaxed.

Facial Feedback Meta-Analysis and Effect Sizes

Nicholas Coles, Jeff Larsen, and Heather Lench (2019) evaluated the evidence for the facial feedback hypothesis with a meta-analysis of 138 research studies that examined and measured the effect size of the facial feedback. Effect size  measures the strength of the relationship between two variables. In this case, the effect size was the impact of the subjects’ mood from the facial manipulation. The authors indicated that the combined effect sizes from the studies they reviewed were small but varied widely.

The authors of this study identified factors that significantly impacted the effect sizes of facial feedback. Larger effect sizes were found for studies that did not include emotional stimulation but relied on manipulated facial muscles. These results suggest that facial musculature movement might have a more significant impact than some emotionally arousing impetus. In addition, facial feedback effect sizes varied depending on the type of stimuli (see graph below). Some stimuli had smaller effect sizes, and some had more significant effect sizes.

The authors concluded that facial feedback could both initiate and modify emotional experiences, and effect size varies based on the emotional stimulation. Studies testing facial feedback using “Imagined Scenarios” showed the most significant response. Imagined scenarios can be similar to the distorted negative cognitions that depressed individuals experience as they “imagine” that their problems are more severe and hopeless than they are.

Questions or comments? We would love to hear from you.

Coles, N. A., Larsen, J. T., & Lench, H. C. (2019). A meta-analysis of the facial feedback literature: Effects of facial feedback on emotional experience are small and variable. Psychological Bulletin , 145 (6), 610–651. https://doi.org/10.1037/bul0000194

Ekman, P., Friesen W.V. (1978).  Facial action coding system: A technique for the measurement of facial movement . Palo Alto: Consulting Psychologists Press.

Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomic Nervous System Activity Distinguishes among Emotions. Science, 221 (4616), 1208–1210. https://doi.org/10.1126/science.6612338

Levenson, R.W., Ekman, P., & Friesen, W. V. (1990). Voluntary facial action generates emotion-specific autonomic nervous system activity. Psychophysiology, 27 , 363–384. https://doi.org/10.1111/j.1469-8986.1990.tb02330.x

Lewis, M.B. (2012). Exploring the positive and negative implications of facial feedback. Emotion , 12 (4), 852–859. https://doi.org/10.1037/a0029275