Most of the research on facial expression has focused on negative emotions, such as anger, fear, or disgust (e.g., Adams et al. Reference Adams, Gordon, Baird, Ambady and Kleck2003; Mathews et al. Reference Mathews, Fox, Yiend and Calder2003; Putman et al. Reference Putman, Hermans and van Honk2006), while positive emotions have received relatively less attention. The attempt to examine the nature of smiles by Niedenthal et al. is therefore highly welcome. To assist with further development of SIMS, we highlight its omission of observer bias and the importance of measuring the time course of motor resonance.
Niedenthal et al.'s proposal as to how a smile may be processed by an observer is rather one-sided. Importantly, they do not take into account the observer's motives and suppositions. As the authors point out, analysis of the facial features may not always be sufficient to classify realistic emotional expressions, and then the perceiver must call on alternative sources of information, such as the social situation in which the expression occurs. More than that, the perceiver will evaluate and interpret the expression on the basis of his or her emotional state and desires. Specifically, if an observer is in a happy mood, then an emotion expressed by another person will be interpreted differently compared to when the observer is sad or angry.
According to the affect-as-information theory (e.g., Schwarz Reference Schwarz, Higgins and Sorrentino1990), mood is used to make social judgments. When a person is in a certain mood, he or she may inadvertently attribute that mood state to other people, which in turn will influence the interpretation of an observed facial expression. In fact, Niedenthal herself (Niedenthal et al. Reference Niedenthal, Halberstadt, Margolin and Innes-Ker2000) found an effect of emotional state in a task where participants had to detect at which frame of a movie a facial expression (e.g., happiness) became neutral. When the participant's mood was congruent with the initially displayed emotion, this expression was perceived to persist longer than when participant mood and initial stimulus expression were incongruent. Hence, a happy person will more readily interpret emotion is not too extreme).a given emotional display as happy (at least as long as the displayed
In the 1950's psychologists argued that what we perceive is influenced by our motivations or “set” (the so-called “New Look”). For example, the size of more valuable coins was overestimated, and hungry participants were quicker than satiated participants to identify food-related words (for critical discussion, see Erdelyi Reference Erdelyi1974). Applying this insight to our point about observer bias, basic motivational tendencies, such as what we wish to see, might also influence the recognition of smiles. Supporting this claim, we recently found a self-referential positivity bias when judging where another person is looking (Lobmaier & Perrett, in press; Lobmaier et al. Reference Lobmaier, Tiddeman and Perrett2008). People expressing a happy emotion were much more likely to be interpreted as attending to the observer than were people with an angry or sad expression. It is beneficial for our self-esteem to assume that we are the reason for somebody else's happiness. Thus, in the context of the SIMS model, expressions may be interpreted as more positive, just because this is what we wish the expression to be: not only will we perceive a smile as directed to us, we might also interpret a facial expression to be more positive than it really is.
The SIMS model explains recognition of smiles as a mainly stimulus-driven process. This approach is in contrast with various current models of emotion recognition. For example, the appraisal model of emotion (e.g., Scherer Reference Scherer2001) postulates that stimuli undergo evaluation checks on the basis of various dimensions: novelty (is a stimulus novel, predictable?), intrinsic pleasantness (positive or negative evaluation), goal significance (is it relevant to pursue my goal?), coping potential (can I deal with it?), and compatibility with social or personal standards (is it acceptable for me and others?). In the context of happy emotions, this means that (unconscious) cognitive evaluation of a seen smile will accompany any simulation of smiles. Such processes are in stark contrast with Niedenthal et al.'s bottom-up view of emotional resonance.
Nevertheless, the SIMS model might be a suitable framework for further exploring the interplay of top-down and bottom-up factors in the categorization of emotions because it raises the issue of strategic influences and their timing. The fact that it takes half a second (Dimberg & Thunberg 1998; cited in sect. 3.2.1 of the target article) or even longer (Bavelas et al. 1986; cited in sect. 4.2.1) before the postulated motor resonance mechanism leads to facial mimicry is noteworthy in the light of much faster motor resonance effects in language comprehension (for a recent review, see Fischer & Zwaan Reference Fischer and Zwaan2008). Contrary to Niedenthal et al.'s claim (in sect. 4), the embodied or simulation view of language comprehension postulates an obligatory simulation process, such that even abstract concepts are understood through metaphorical grounding (Barsalou Reference Barsalou2008). Further chronometric studies of comprehension processes that combine methods from different fields are needed to further refine the proposed resonance mechanism.
We argue that a model attempting to account for recognition of smiles is only complete if stimulus-driven bottom-up processes are integrated with endogenous top-down processes. Indeed, the happy expression will be simulated, but we might expect that both the rapid simulation and the subsequent interpretation of the smile will be monitored by more frontal regions such as the medial prefrontal cortex (cf. Keysers & Gazzola Reference Keysers and Gazzola2006). To classify a smile only by simulation would assume that a certain smile will mean the same thing in every social situation. This is not the case. In some situations, we, as observers, will reflect on the meaning of the smile just as we reflect on our own state and will consequently interpret the meaning of the smile based on more than just simulation. Thus, interpreting a smile goes far beyond using simulation alone.
You have
Access
Most of the research on facial expression has focused on negative emotions, such as anger, fear, or disgust (e.g., Adams et al. Reference Adams, Gordon, Baird, Ambady and Kleck2003; Mathews et al. Reference Mathews, Fox, Yiend and Calder2003; Putman et al. Reference Putman, Hermans and van Honk2006), while positive emotions have received relatively less attention. The attempt to examine the nature of smiles by Niedenthal et al. is therefore highly welcome. To assist with further development of SIMS, we highlight its omission of observer bias and the importance of measuring the time course of motor resonance.
Niedenthal et al.'s proposal as to how a smile may be processed by an observer is rather one-sided. Importantly, they do not take into account the observer's motives and suppositions. As the authors point out, analysis of the facial features may not always be sufficient to classify realistic emotional expressions, and then the perceiver must call on alternative sources of information, such as the social situation in which the expression occurs. More than that, the perceiver will evaluate and interpret the expression on the basis of his or her emotional state and desires. Specifically, if an observer is in a happy mood, then an emotion expressed by another person will be interpreted differently compared to when the observer is sad or angry.
According to the affect-as-information theory (e.g., Schwarz Reference Schwarz, Higgins and Sorrentino1990), mood is used to make social judgments. When a person is in a certain mood, he or she may inadvertently attribute that mood state to other people, which in turn will influence the interpretation of an observed facial expression. In fact, Niedenthal herself (Niedenthal et al. Reference Niedenthal, Halberstadt, Margolin and Innes-Ker2000) found an effect of emotional state in a task where participants had to detect at which frame of a movie a facial expression (e.g., happiness) became neutral. When the participant's mood was congruent with the initially displayed emotion, this expression was perceived to persist longer than when participant mood and initial stimulus expression were incongruent. Hence, a happy person will more readily interpret emotion is not too extreme).a given emotional display as happy (at least as long as the displayed
In the 1950's psychologists argued that what we perceive is influenced by our motivations or “set” (the so-called “New Look”). For example, the size of more valuable coins was overestimated, and hungry participants were quicker than satiated participants to identify food-related words (for critical discussion, see Erdelyi Reference Erdelyi1974). Applying this insight to our point about observer bias, basic motivational tendencies, such as what we wish to see, might also influence the recognition of smiles. Supporting this claim, we recently found a self-referential positivity bias when judging where another person is looking (Lobmaier & Perrett, in press; Lobmaier et al. Reference Lobmaier, Tiddeman and Perrett2008). People expressing a happy emotion were much more likely to be interpreted as attending to the observer than were people with an angry or sad expression. It is beneficial for our self-esteem to assume that we are the reason for somebody else's happiness. Thus, in the context of the SIMS model, expressions may be interpreted as more positive, just because this is what we wish the expression to be: not only will we perceive a smile as directed to us, we might also interpret a facial expression to be more positive than it really is.
The SIMS model explains recognition of smiles as a mainly stimulus-driven process. This approach is in contrast with various current models of emotion recognition. For example, the appraisal model of emotion (e.g., Scherer Reference Scherer2001) postulates that stimuli undergo evaluation checks on the basis of various dimensions: novelty (is a stimulus novel, predictable?), intrinsic pleasantness (positive or negative evaluation), goal significance (is it relevant to pursue my goal?), coping potential (can I deal with it?), and compatibility with social or personal standards (is it acceptable for me and others?). In the context of happy emotions, this means that (unconscious) cognitive evaluation of a seen smile will accompany any simulation of smiles. Such processes are in stark contrast with Niedenthal et al.'s bottom-up view of emotional resonance.
Nevertheless, the SIMS model might be a suitable framework for further exploring the interplay of top-down and bottom-up factors in the categorization of emotions because it raises the issue of strategic influences and their timing. The fact that it takes half a second (Dimberg & Thunberg 1998; cited in sect. 3.2.1 of the target article) or even longer (Bavelas et al. 1986; cited in sect. 4.2.1) before the postulated motor resonance mechanism leads to facial mimicry is noteworthy in the light of much faster motor resonance effects in language comprehension (for a recent review, see Fischer & Zwaan Reference Fischer and Zwaan2008). Contrary to Niedenthal et al.'s claim (in sect. 4), the embodied or simulation view of language comprehension postulates an obligatory simulation process, such that even abstract concepts are understood through metaphorical grounding (Barsalou Reference Barsalou2008). Further chronometric studies of comprehension processes that combine methods from different fields are needed to further refine the proposed resonance mechanism.
We argue that a model attempting to account for recognition of smiles is only complete if stimulus-driven bottom-up processes are integrated with endogenous top-down processes. Indeed, the happy expression will be simulated, but we might expect that both the rapid simulation and the subsequent interpretation of the smile will be monitored by more frontal regions such as the medial prefrontal cortex (cf. Keysers & Gazzola Reference Keysers and Gazzola2006). To classify a smile only by simulation would assume that a certain smile will mean the same thing in every social situation. This is not the case. In some situations, we, as observers, will reflect on the meaning of the smile just as we reflect on our own state and will consequently interpret the meaning of the smile based on more than just simulation. Thus, interpreting a smile goes far beyond using simulation alone.