According to the Simulation of Smiles (SIMS) model, simulation allows one to discriminate between different types of smiles. Some types of smiles (such as “true” and “false” smiles) are associated with particular physical characteristics (Cacioppo et al. Reference Cacioppo, Petty, Losch and Kim1986; Ekman et al. Reference Ekman, Friesen and Ancoli1980 – reviewed in Frank Reference Frank and Abel2002; see also, Hess & Kleck Reference Hess and Kleck1990). Other types of smiles (e.g., affiliative vs. enjoyment smiles) cannot be discriminated from each other on the basis of physical markers. In this case, the meaning of a smile is derived from semantic/contextual information. Because smile types can be identified based on either physical or contextual cues, this suggests that any information provided by simulation simply serves to enhance the identification of the smile's meaning (e.g., by speeding up the recognition process), rather than being critical in recognizing the smile.
One study which suggests that simulation may be critical in recognizing smiles was conducted by Maringer et al. (in press; cited in the target article). This study showed that mimicry can facilitate the discrimination of true from false dynamic smiles. However, this advantage may be due to the fact that mimicry provides an additional representation of a smile stimulus, which allows participants who mimic to receive more input about a smile than participants who do not mimic. Future experiments can control for the number of stimulus representations between the mimicry and no-mimicry conditions by adopting a yoked design. This would involve videotaping the faces of participants in the mimicry condition when they view a sequence of face stimuli, and playing these recordings to yoked participants in the no-mimicry condition while they view the same stimulus sequence. In this way, any differences between the two conditions can be attributed to qualitative differences in stimulus processing, rather than to differences in the number of stimulus representations received.
According to the SIMS model, smiles elicited when viewing positive (i.e., enjoyment) smiles indicate mimicry. In this case, it is the current expression and feeling of the expresser that is simulated. Niedenthal et al. argue that, by contrast, people are less likely to mimic negative (e.g., dominant) smiles, but instead, they simulate the feeling of being dominated. In this case, it is the previous experience of the perceiver that is being simulated. It is not clear why the different smiles should invoke simulations with reference to different people- and time- perspectives. Instead, it would be more parsimonious to claim that the same mechanism (i.e., mimicry) occurs for both types of smiles, at least when the smiles are presented in the same context, such as when viewing smile stimuli in a laboratory setting. In support of the idea that dominant smiles may be mimicked, studies show that people spontaneously mimic another dominant expression, anger, when viewing pictures of angry faces (e.g., Cannon et al. Reference Cannon, Hayes and Tipper2009; Dimberg et al. Reference Dimberg, Thunberg and Elmehed2000). The question of whether mimicry occurs similarly for both positive and negative smiles may be answered by analyzing the results of Maringer et al. (in press) in more detail. If Niedenthal et al. consider that people should inhibit mimicry for false smiles as they would for dominant smiles (because, to extend the authors' reasoning, it is not socially appropriate to mimic false smiles), then people should not mimic false smiles in the no-mimicry condition of Maringer et al.'s study, but they should nevertheless be able to simulate reactions that they have previously felt in response to false smiles. As such, participants in the mimicry and no-mimicry conditions should recognize false smiles with equal accuracy. However, if participants spontaneously mimic false smiles, then false smiles should be recognized more accurately in the mimicry than in the no-mimicry condition.
Niedenthal et al. claim that eye contact increases emotional intensity in response to angry and happy faces (Schrammel et al. Reference Schrammel, Pannasch, Graupner, Mojzisch and Velichkovsky2009) because it triggers the simulation of expressions and their associated emotions. However, the relation between eye contact and emotional intensity may be explained by other mechanisms. Eye contact may enhance the emotional experience of the perceiver because it implies that the perceiver is the target of the expression, which in turn increases the relevance of the expressed emotion for the perceiver (e.g., Cary Reference Cary1978). Alternatively, emotional intensity may increase under eye contact because angry and happy faces both signal an approach motivation, which is congruent with the approach motivation signaled by a direct gaze, but incongruent with the avoidance motivation signaled by an averted gaze (see Argyle & Cook Reference Argyle and Cook1976; Davidson & Hugdahl Reference Davidson and Hugdahl1995; Harmon-Jones & Segilman Reference Harmon-Jones and Segilman2001). Thus, happy faces with an averted gaze may be considered to be more ambiguous than happy faces with a direct gaze, causing them to elicit less emotion, or ambivalent emotion. In support of this, studies show that when the approach/avoidance motivation of the gaze direction matches the approach/avoidance motivation of the emotion expression (e.g., when happy and angry faces are presented with a direct gaze, and sad and fearful faces are presented with an averted gaze), the emotion is identified more quickly (Adams & Kleck Reference Adams and Kleck2003; Reference Adams and Kleck2005), and the perceived emotional intensity of the expression also increases (Adams & Kleck Reference Adams and Kleck2005). These results with happy and angry faces are consistent with the SIMS model. However, the model assumes that similar processes should underlie the simulation of all types of expressions, and thus it cannot account for why sad and fearful faces should be identified more quickly, and considered to be more emotionally intense, when presented with averted rather than with direct gazes.
The SIMS model represents an ambitious integration of social, cognitive, and neuro-psychology. However, one aspect that needs to be clarified is whether simulation operates similarly for different types of smiles, such as true versus false smiles, and smiles associated with physical markers versus more ambiguous smiles. It also remains to be demonstrated whether simulation is indeed critical for the interpretation of smiles, or whether physical and contextual cues play a greater role in this process. Nevertheless, the model raises many testable hypotheses, which provide a fruitful starting point for exploring the contribution of simulation to the processing of facial expressions.
According to the Simulation of Smiles (SIMS) model, simulation allows one to discriminate between different types of smiles. Some types of smiles (such as “true” and “false” smiles) are associated with particular physical characteristics (Cacioppo et al. Reference Cacioppo, Petty, Losch and Kim1986; Ekman et al. Reference Ekman, Friesen and Ancoli1980 – reviewed in Frank Reference Frank and Abel2002; see also, Hess & Kleck Reference Hess and Kleck1990). Other types of smiles (e.g., affiliative vs. enjoyment smiles) cannot be discriminated from each other on the basis of physical markers. In this case, the meaning of a smile is derived from semantic/contextual information. Because smile types can be identified based on either physical or contextual cues, this suggests that any information provided by simulation simply serves to enhance the identification of the smile's meaning (e.g., by speeding up the recognition process), rather than being critical in recognizing the smile.
One study which suggests that simulation may be critical in recognizing smiles was conducted by Maringer et al. (in press; cited in the target article). This study showed that mimicry can facilitate the discrimination of true from false dynamic smiles. However, this advantage may be due to the fact that mimicry provides an additional representation of a smile stimulus, which allows participants who mimic to receive more input about a smile than participants who do not mimic. Future experiments can control for the number of stimulus representations between the mimicry and no-mimicry conditions by adopting a yoked design. This would involve videotaping the faces of participants in the mimicry condition when they view a sequence of face stimuli, and playing these recordings to yoked participants in the no-mimicry condition while they view the same stimulus sequence. In this way, any differences between the two conditions can be attributed to qualitative differences in stimulus processing, rather than to differences in the number of stimulus representations received.
According to the SIMS model, smiles elicited when viewing positive (i.e., enjoyment) smiles indicate mimicry. In this case, it is the current expression and feeling of the expresser that is simulated. Niedenthal et al. argue that, by contrast, people are less likely to mimic negative (e.g., dominant) smiles, but instead, they simulate the feeling of being dominated. In this case, it is the previous experience of the perceiver that is being simulated. It is not clear why the different smiles should invoke simulations with reference to different people- and time- perspectives. Instead, it would be more parsimonious to claim that the same mechanism (i.e., mimicry) occurs for both types of smiles, at least when the smiles are presented in the same context, such as when viewing smile stimuli in a laboratory setting. In support of the idea that dominant smiles may be mimicked, studies show that people spontaneously mimic another dominant expression, anger, when viewing pictures of angry faces (e.g., Cannon et al. Reference Cannon, Hayes and Tipper2009; Dimberg et al. Reference Dimberg, Thunberg and Elmehed2000). The question of whether mimicry occurs similarly for both positive and negative smiles may be answered by analyzing the results of Maringer et al. (in press) in more detail. If Niedenthal et al. consider that people should inhibit mimicry for false smiles as they would for dominant smiles (because, to extend the authors' reasoning, it is not socially appropriate to mimic false smiles), then people should not mimic false smiles in the no-mimicry condition of Maringer et al.'s study, but they should nevertheless be able to simulate reactions that they have previously felt in response to false smiles. As such, participants in the mimicry and no-mimicry conditions should recognize false smiles with equal accuracy. However, if participants spontaneously mimic false smiles, then false smiles should be recognized more accurately in the mimicry than in the no-mimicry condition.
Niedenthal et al. claim that eye contact increases emotional intensity in response to angry and happy faces (Schrammel et al. Reference Schrammel, Pannasch, Graupner, Mojzisch and Velichkovsky2009) because it triggers the simulation of expressions and their associated emotions. However, the relation between eye contact and emotional intensity may be explained by other mechanisms. Eye contact may enhance the emotional experience of the perceiver because it implies that the perceiver is the target of the expression, which in turn increases the relevance of the expressed emotion for the perceiver (e.g., Cary Reference Cary1978). Alternatively, emotional intensity may increase under eye contact because angry and happy faces both signal an approach motivation, which is congruent with the approach motivation signaled by a direct gaze, but incongruent with the avoidance motivation signaled by an averted gaze (see Argyle & Cook Reference Argyle and Cook1976; Davidson & Hugdahl Reference Davidson and Hugdahl1995; Harmon-Jones & Segilman Reference Harmon-Jones and Segilman2001). Thus, happy faces with an averted gaze may be considered to be more ambiguous than happy faces with a direct gaze, causing them to elicit less emotion, or ambivalent emotion. In support of this, studies show that when the approach/avoidance motivation of the gaze direction matches the approach/avoidance motivation of the emotion expression (e.g., when happy and angry faces are presented with a direct gaze, and sad and fearful faces are presented with an averted gaze), the emotion is identified more quickly (Adams & Kleck Reference Adams and Kleck2003; Reference Adams and Kleck2005), and the perceived emotional intensity of the expression also increases (Adams & Kleck Reference Adams and Kleck2005). These results with happy and angry faces are consistent with the SIMS model. However, the model assumes that similar processes should underlie the simulation of all types of expressions, and thus it cannot account for why sad and fearful faces should be identified more quickly, and considered to be more emotionally intense, when presented with averted rather than with direct gazes.
The SIMS model represents an ambitious integration of social, cognitive, and neuro-psychology. However, one aspect that needs to be clarified is whether simulation operates similarly for different types of smiles, such as true versus false smiles, and smiles associated with physical markers versus more ambiguous smiles. It also remains to be demonstrated whether simulation is indeed critical for the interpretation of smiles, or whether physical and contextual cues play a greater role in this process. Nevertheless, the model raises many testable hypotheses, which provide a fruitful starting point for exploring the contribution of simulation to the processing of facial expressions.
ACKNOWLEDGMENT
The writing of this commentary was facilitated by a grant from the “Action de Recherche Concertée”; No. ARC06/11-337.