The existence of a 2-way connection between facial expression and emotion has been an enduring concept; however, the connection’s directional bias and strength have been points of enduring contention. The name used to describe this relationship is the Facial Feedback Hypothesis. In scientific literature it was initially presented in two different forms, with both origins from other a century ago. The first version relates to Darwin’s (1872) suggestion that:
The free expression by outward signs of an emotion intensifies it. On the other hand, the repression, as far as this is possible, of all outward signs softens our emotions… Even the simulation of an emotion tends to arouse it in our minds. (p. 336)
Best services for writing your paper according to Trustpilot
In summation this version suggests that the strength of an emotion can be modulated by expression, and perhaps an emotion can be initiated by expression. Though this itself later split into two variants.
The second version’s origins could be best placed with the beliefs of William James. James (1884) philosophised that the only emotions worth considering as real were those associated with a physical expression. Furthermore he hypothesised that it was the act of making an outward expression which caused the internal feeling of an emotion, and that this expression was necessary for emotion initiation. That is to say he suggested people are only afraid because they tremble (James, 1884) rather than trembling because they are afraid. It seems James may have blurred the lines between emotional responses like fear or happiness with physical processes like pain or hunger in his hypothesis of the physiological initiation of emotion.
Thus the Facial Feedback Hypothesis developed more from Darwin’s beliefs rather than James’, even if the definitions within this century varied considerably from article to article.
In the late 1980’s Adelmann and Zajonc reviewed the work on the subject by their contemporaries’, summarising that “facial efference plays not only a modulating function but an initiating function in the experience of emotion” (Adelmann & Zajonc, 1989). Though they finish with a more restrained conclusion of “facial efference under some conditions may modulate or even initiate subjective emotional experience” (Adelmann & Zajonc, 1989). What is notable is the suggestion that facial feedback, or facial efference as they call it, can initiate emotional experience. With modulation of emotional experience the suggestion is that facial muscle contractions or inhibitions associated with emotions can change the intensity of already occurring emotions. An example could be disallowing a smile when happy, could lead to a less enjoyable experience being had. It would still be a positive experience, just not to the same intensity that it would be if the natural smile was able to be expressed (Bush, Barr, McHugo, & Lanzetta, 1989). This has come to be known as the weak version of facial feedback hypothesis (L. L. Rutledge & Hupka, 1985).
In comparison the initiation of emotional experience is the suggestion that creating an expression through facial muscle contractions will lead to the creation of an emotional experience. An example could be to furrow the brow (Larsen, Kasimatis, & Frey, 1992), with the result being that the person started to feel slightly sad when they had previously been neutral. This has come to be known as the strong version of facial feedback hypothesis (L. L. Rutledge & Hupka, 1985). Both the strong and weak hypothesises can be traced back to Darwin (1872).
The landmark experimental test of facial feedback hypothesis had participants unwittingly contract their facial muscles to simulate part of a facial expression. Strack, Martin and Stepper (1988) had participants hold pencils in their mouths in various ways to manipulate the expression made by the lower half of their face. By holding the pencil tip in their mouth between either their teeth or their lips related respectively to smiling or frowning. The reason that this experiment stood out was because it managed to get participants to conform to making the partial expressions without them being aware they were doing, reducing confounders. In most previous successful experiments participants knew they were simulating or resisting expressions of happiness, sadness, shock or disgust (Colby, Lanzetta, & Kleck, 1977; Cupchik & Leventhal, 1974; Leventhal & Mace, 1970).
Though many experiments continued to use procedures that involved explicitly instructing participants to make expressions (Dimberg & Soderkvist, 2010; Kleinke, Peterson, & T. Rutledge, 1998; Strack & Neumann, 2000), it gave rise to experiments that could study the facial movements and inhibitions in a number of subliminal ways. These implicit methods included cover stories for movements where participants make the expressions unwittingly (Larsen et al., 1992), or like Davis, Senghas and Ochsner (2009) inhibition condition where they applied electrodes to participants’ faces and told them if they moved their faces, it would ruin the readings they were taking. External mechanisms have also been used to cause the expression without requirement for the participant to move their muscles actively (Mori & Mori, 2009). Even chemical interaction conditions have come into use.
Understandably in some experiments that try to implicitly manipulate expressions, participants realise they are being made to make an expression. In Reisenzein and Studtmann’s (2007) experiment on the shock expression they had participants look at a large screen above them causing them to open their eyes wider. At the end of the experiement many reported they felt they were mimicking a hock expression, but blamed part of their self awareness of this on the fact participants had been prompted into a shock expression before being placed into their condition. Also, though it is not brought into their discussion or critique in the mechanism based smiling experiment of Mori & Mori (2009), it is suspected that many participants realised that the act of smiling was being simulated as the adhesive bandages applied to their cheeks were being pulled upwards by elastic bands.
However, an experimental method which can be used to test facial feedback hypothesis as a reversible lesion model which is absent of participant awareness has came to light. That is the use of Botulinum Toxin, known commercially as Botox, to reduce or restrict movement of isolated muscle groups (Davis, Senghas, Brandt, & Ochsner, 2010; Havas, Glenberg, Gutowski, Lucarelli, & Davidson, 2010; Lewis & Bowler, 2009). Methodologically speaking it would be very difficult to obtain ethical approval and participant consent to use Botox on participants’ facial muscles without explanation of why an experimenter wishes to do so. That is why many experimenters in this topic have started to recruit participants from beauty centres and cosmetic surgeries where those that would become the participants already had the intention of using Botox (Alam, Barrett, Hodapp, & Arndt, 2008; Davis et al., 2010; Lewis & Bowler, 2009).
It is important to explain how the effect of facial feedback hypothesis is measured. To be critical, there is no single way to measure the effect of facial feedback hypothesis, only whatever measurement an experimenter can justify has something to do with emotion. Some methods have stronger backing than others, but can vary due to funding. One such way proposed to measure the effect is to have participants rate the experience of presented stimuli on a scale of emotionality. Strack et al. (1988) did this by having participants rate how funny they thought cartoon strips were, comparing their conditions of having a pencil held between the lips like a frown, between the teeth like a smile or in the non-dominant hand as a control.
Some experiments have participants choose which emotion they feel more aligned to, using closed questions; others use emotionality scales that move from happy to sad or angry to calm (Mori & Mori, 2009). In a neuroscience view to showing an effect Hennenlotter et al. (2009) used Functional Magnetic Resonance Imaging (fMRI) to show attenuation in the activation of the left amygdala, a key component in emotion processing, when their participants had undergone Botox treatment to their corrugator muscle reducing the intensity of a frown. It has even been shown that the involuntary movements of the face are causally related to reading times of emotionally congruent sentences. That is to say smiling slightly can make someone faster at reading a happy or positively orientated piece of text. This effect was proven by Havas et al. (2010) when they used Botox to reduce facial expressiveness in participants who then had to read emotive pieces of text. In a previous experiment by Havas, Glenberg and Rinck (2007) they showed more simply that the speed of reading emotional sentences could be modulated using the pencil in the mouth facial feedback method created by Strack et al. (1988).
The Present Study
In the present study the primary aim is to compare the effect of self awareness in expression making on the effectiveness of the facial feedback hypothesis. That is conditions where the participants know they are making or inhibiting expressions will be compared against conditions where they are unaware they are doing so.
The aspect of the facial muscle system this study is interested in is the corrugator supercilious. This muscle is responsible for drawing the eyebrows downward and towards each other, creating a main component of a frown. To investigate the movement of this facial muscle without having to use a contrived cover story for generic participants, it is the experimenter’s intention to gather participants from those seeking Botox treatment for their fore brow. Botox neurotoxin causes temporary muscular denervation and is used in cosmetic treatment efforts to reduce the signs of expression which come with aging (Fagien, Rohrich, & Janis, 2003).
To investigate the facial feedback involved with the corrugator muscle 4 conditions will be used; two conditions of movement or expression, and 2 conditions of restraint. In the unaware-inhibit condition patients that have recently underwent subcutaneous Botox treatment to the corrugator muscle will be recruited as their corrugator muscle will be less able to contract due to treatment. The other three conditions will use pre-Botox patients as participants. The aware-inhibit condition will have participants asked to actively resist making negative facial expressions with the upper part of their face, examples being demonstrated if required. The aware-expression condition will ask patients to actively hold their eyebrows like a frown for the experiment. Finally the unaware-expression condition will have participants instructed to taughten their forehead, with guidance to lower their eyebrows if they do not do this at first attempt.
The stimulus and measurement used in the experiment will be similar to that of Havas et al. (2010). It includes having the participants read emotionally positive or negative sentences which appear on a computer screen in a random order. Once a sentence appears participants are to read it and press a key to confirm they understand the sentence. It is the time interval between the sentence appearing and the participant confirming their understanding which will be recorded for analysis. Reading time aggregates will be compared within participants for the types of sentence; negative and positive. The aggregates will be compared between conditions. The experimental method will vary from Havas et al.’s in that only one session of sentence reading will take place as that experiment wished to test before and after Botox treatment whereas this experiment wishes to test varying conditions.
Hypothesises
It is hypothesised based on previous research that the expressive groups (aware and unaware) will have faster reading times for negative sentences than the inhibition groups (aware and unaware)
It is hypothesised based on previous research that the expressive groups will have slower reading times for the positive sentences than the inhibition groups
It is hypothesised newly that the unaware-expression condition will have reading times faster for negative sentences and slower for positive sentences than the aware-expression condition.
It is hypothesised newly that the unaware-inhibit (Botox) condition will have reading times slower for negative sentences and equal for positive sentences than the aware-inhibit condition.
All hypothesises are summarised in Matrix 1.
Negative Sentences
Positive Sentences
Slowest —– Fastest
unaware-expression
Slowest —– Fastest
unaware-inhibit
aware-inhibit
aware-expression
aware-inhibit
aware-expression
unaware-inhibit
Unaware-expression
Matrix 1: Predicted ordering of conditions’ reading times for negative and positive sentences
METHOD
Pilot
To carry out this experiment, a series of similar length sentences will need to be designed; approximately 40 sad and 40 happy sentences. Pilot participants will recruited for course credit and will need to rate each sentence on levels of sadness and happiness. Sentences strong in one attribute and weak in the other should be kept. While sentences that rate as more ambivalent should be discarded. Examples of sentences used in another study can be seen in Table 1.
Sentence type
Example Sentences
Happy
The water park is refreshing on the hot summer day.
Finally, you reach the summit of the tall mountain.
You spring up the stairs to your lover’s apartment.
Sad
You hold back your tears as you enter the funeral home.
You open your email in-box on your birthday to find no new emails.
Your closest friend has just been hospitalized for a mental illness
Table 1. Example emotional sentences (Havas et al., 2010)
Participants
Sixty participants will be recruited from would be patients wanting to receive Botox treatment to their corrugator supercilious in order to reduce frown lines and wrinkles. Participants will be recruited through the cosmetic surgeries themselves with the offer that if they take part in the experiment a sum of money will be provided to help offset the cost of treatment. The participants will be randomly assigned to their condition. Those in the unaware-inhibit condition will go through the experiment session post treatment, whereas all other conditions will take part prior to treatment.
Materials
An experiment software program which can record reaction times will be required. It needs to be able to prime sentences and record the time between initial priming and reaction. A computer or laptop will also be required to display and carry out the program. Twenty of each of the best happy and sad sentences from the pilot process should be brought forward to this experiment and inputted into the experiment software program.
Design
The experiment has 2 independent variables. The first is the between participants condition which they are randomly assigned to. The 4 conditions are unaware-inhibit, aware-inhibit, aware-expression and unaware-expression. The second independent variable is within participants and the conditions are the two types of sentences that will be shown to participants; sad and happy. The dependent variable in this experiment is the recorded reading times of each participant for the emotional sentences read and will be measured in milliseconds (ms).
Procedure
