Understanding Sparks of Emotion

Facial expressions of emotion—such as the joyful smile you might display when encountering a friend or your angry frown when being cut off in traffic—are powerful social signals that are able to evoke emotional reactions in others. Emotional expressions, however, can also profoundly influence the expresser, changing the way they feel and perceive the world.

This bidirectional relationship between emotional expression and experience was first hypothesized by Charles Darwin. The facial feedback hypothesis (FFH) was formalized in the 20^th century. It proposes that activating one’s facial muscles can generate new feelings or modulate one’s ongoing emotions, as well as change the way others’ emotions are perceived. Indeed, hindering people from spontaneously mimicking others’ emotional expressions (called “facial mimicry,” something we all tend to do, believe it or not) also diminishes their ability to recognize them. The reason is that information about the state of muscle activation or relaxation is constantly fed back from the face to the brain (Wood et al., 2016).

To test these ideas scientifically, various methods have been used to either activate facial muscles—and thus generate the corresponding emotion—or instead inhibit them and, therefore, prevent or reduce spontaneous expression, including facial mimicry. For example, participants report feeling happier when posing a smile and less happy when posing a frown (Coles et al., 2022; Laird, 1974). Expressing an emotion in the face can also change how images or videos are perceived. For example, cartoons are perceived as more humorous when smiling than when frowning (Laird, 1974).

Another often-used method consists of asking participants to hold a pen or chopstick either sideways between the teeth, without touching it with the lips (which creates a smile-like expression), or lengthwise between the lips without using the teeth (which makes smiling nearly impossible). Using the pen-in-mouth technique, it was found, for example, that smiling, compared to frowning, makes cartoons appear funnier (Strack et al., 1988) and lowers the threshold of perceiving happiness in faces and bodies (Marmolejo-Ramos et al., 2020).

Further important methods involve studying individuals who have some form of facial muscle paralysis, be this desired and short-lived, as after Botox injections (Davis et al., 2010), unwanted and longer-lasting, e.g., due to Bell’s palsy (Korb et al., 2016), or even congenital due to Moebius syndrome or other forms of inborn facial paralysis (Schiano Lomoriello et al., 2023).

Using the techniques mentioned above, as well as other similar ones, scholars have amassed considerable evidence supporting the FFH’s claim of an interaction between one’s facial muscle activity, on the one hand, and one’s felt emotion and perception of the world on the other hand (Wood et al., 2016). This work is relevant not just from a scientific point of view but also because it can help improve well-being. Indeed, “laughter yoga” has been inspired by it, and research in this domain may help improve emotion recognition in groups with social interaction difficulties, such as autism.

Nevertheless, the FFH has remained highly controversial, as its related effects tend to be weak and variable (Coles et al., 2019). For example, the seminal finding from the 1980s that holding a pen between the teeth, thereby mimicking a smile, increases feelings of amusement toward cartoons was not replicated in a recent multi-lab study (Wagenmakers et al., 2016). A major cause for these inconsistencies in the literature might be the limitations of the methods employed to test the FFH.

Specifically, current techniques typically do not allow researchers to precisely control which facial muscles are (de)activated, how strongly and symmetrically the muscle contraction or inhibition is, or how long it will last. For example, voluntary posing of facial expressions, as well as holding a pen only with the lips, is effortful and cannot be adequately achieved by all participants. Furthermore, research with people who had Botox injections or suffer from muscle paralysis can be complicated because those effects last for at least several months, and thus compensatory mechanisms can arise.

Duchenne de Boulogne applies electrical stimulation to facial muscles

Source: Wikipedia / Public Domain

To overcome the limitations of previously used methods to activate or inhibit facial muscles, allow for finer control of facial muscle activation (in space, time, and intensity), and propose an alternative way to test aspects of the FFH, my laboratory has recently started using electrical impulses to stimulate facial muscles. This technique, called facial neuromuscular electrical stimulation (fNMES, see here for a video demonstration), consists of placing electrodes on the skin and delivering short trains of weak electrical pulses to stimulate the underlying muscles. It dates back to the 19th-century neurologist Duchenne de Boulogne, whose photographs of electrically induced facial expressions also inspired Charles Darwin’s work on emotion.

Emotions Essential Reads

How to Settle Painful Feelings for the Long Term

The Guiding Light of Emotions

Despite its long history, fNMES remains largely unknown to modern-day psychologists, which is a shame given the formidable level of control it can provide, especially when being computer controlled. To change this, we have published a set of guidelines on how to conduct fNMES in psychophysiological research (Efthimiou, Hernandez, et al., 2023), and conducted a series of experiments to test aspects of the FFH.

In a recently published experiment (Efthimiou et al., 2024), we showed faces displaying faint happiness or sadness to 47 participants, who decided which emotion was expressed. In half of the trials, fNMES was concurrently applied for half a second to the main smiling muscle. Results showed that fNMES-induced smiling increases the likelihood that faces were perceived as happy compared to trials without stimulation. This finding indicates that smiling can bias our perception of others’ facial expressions, which is in line with the FFH.

In a second experiment (see preprint: Efthimiou, Baker, et al., 2023), we asked if fNMES can change how people feel themselves. A group of 58 participants was asked to report how positively or negatively they felt after receiving five seconds of fNMES to either their smiling or frowning muscles. As expected, participants felt more positive after smiling than frowning, even controlling for the level of discomfort that can sometimes be experienced with fNMES. This result supports the other component of the FFH, i.e., that expressing facial emotions can modulate our own feelings.

It is important to emphasize that these results were obtained even though only low electrical currents were delivered, which resulted in weak activations of individual muscles. Thus, smiling and frowning were only partially re-created—and at weak intensities. Instead, previous experiments supporting the FFH entailed stronger expressions, where several facial muscles were activated to a greater degree (e.g., sincere, strong smiles also include “crow’s feet” around the eyes). Moreover, we found these results despite the fact that fNMES was only delivered for very short durations—as little as half a second in the first of the experiments described above—while previous experiments assessed the effects of prolonged muscle activation or inhibition. We can thus theorize that even stronger effects might be obtained if fNMES were applied with stronger currents and for longer durations—although that comes with the greater risk of inducing discomfort.

We are now planning and conducting further experiments to investigate the optimal parameters for fNMES. For example, we aim to vary the time at which fNMES and face stimuli are presented and also investigate the neural correlates of the emotion perception bias reported above. We have indeed shown that fNMES can be combined with electroencephalography (Baker et al., 2023), allowing us to assess with millisecond precision which brain areas show changes in neuronal activity in response to specific facial muscle activation. Another important aim is to assess if fNMES may be beneficial to improve mood and emotion recognition in individuals with Parkinson’s disease or autism spectrum conditions.

In summary, evidence suggests that electrical stimulation of facial muscles allows us to modulate how people feel and perceive emotions in others. By allowing us to finely control which muscles are activated when and to what degree, fNMES is a powerful tool for science (e.g., to test aspects of the FFH), but importantly, it may also have potential for clinical applications.