One problem for the Simulation of Smiles Model (SIMS) proposed by Niendenthal et al. is that there are individual differences in face-based emotion recognition (hereafter emotion recognition), mimicry, and embodied cognition. Given that these are important features of the model, the model's predictive power may vary as a function of these features. Sex is one important individual difference that needs to be considered (Cahill Reference Cahill2006; Vigil Reference Vigil2009).

Though facial mimicry appears to improve emotion recognition (Niendenthal et al. 2001; Oberman et al. Reference Oberman, Winkielman and Ramachandran2007; Stel & van Knippenberg Reference Stel and van Knippenberg2008) – consistent with the SIMS model – the link between mimicry and emotion recognition has only been reported in females. This could be because women, compared to men, rely more on facial feedback for emotion recognition, at least for some expressions. That is, when facial mimicry is inhibited, women's emotion recognition suffers, while men's does not (Stel & Knippenberg Reference Stel and van Knippenberg2008). Whereas women appear to rely more on this fast and efficient “hot” (embodied) simulatory route for emotion recognition, men may rely more on one or more “cold” (rule-based, non-embodied) strategies. This may account for the finding that women are better than men at distinguishing a number of facial expressions, including smiles (Krumhuber et al. Reference Krumhuber, Manstead and Kappas2007).

Behavioral and neuroimaging evidence is congruent with the notion of sex differences in emotion recognition strategies. For example, men and women have different face viewing patterns during emotion recognition (Hall et al. Reference Hall, Hutton and Morgan2010). While women look more at the eyes, compared to men, men have longer first fixations to the eyes and mouth, perhaps reflecting men's slower processing speed for information contained in those areas (Hall et al. Reference Hall, Hutton and Morgan2010). Additionally, there are anatomical differences between men and women in brain regions containing mirror neurons, which underlie simulation during emotion recognition (Goldman & Sripada Reference Goldman and Sripada2005). During empathetic face-to-face interactions, women, compared to men, experience more activation in brain regions containing mirror neurons (Schulte-Rüther et al. Reference Schulte-Rüther, Markowtsch, Shah, Fink and Piefke2008). Moreover, brain regions containing mirror neurons differ in men and women, both structurally and functionally (Cheng et al. Reference Cheng, Chou, Decety, Chen, Hung, Tzeng and Lin2009).

Researchers examining sex differences in emotion recognition should use caution, for the complexity of measuring emotion recognition may not be apparent. For example, stimulus materials should be selected carefully because women are more sensitive to dynamic, compared to static expressions, while men are not (Biele & Grabowska Reference Biele and Grabowska2006). The gender of face stimuli also influences the perceived emotion: both men and women are better at detecting negative expressions on male faces and positive expressions on female faces (Becker et al. Reference Becker, Kenrick, Neuberg, Blackwell and Smith2007). Also, males and females are not mimicked to the same degree: females are mimicked more often than males (e.g., Lakin et al. Reference Lakin, Chartrand and Arkin2008).

To complicate things further, the likelihood of mimicking depends on top-down influences, including one's conscious or unconscious goals, such as to recognize a face, to attract a potential mate, to understand a spoken language, or to affiliate (e.g., after social rejection). These different motivations alter our viewing patterns of faces and the feature to which we attend (e.g., Saether et al. Reference Saether, Van Belle, Laeng, Brennen and Øvervoll2009; Schyns et al. Reference Schyns, Bonnar and Gosselin2002). If we are attending to a non-emotional aspect of a face – such as skin color – we mimic less (Cannon et al. Reference Cannon, Hayes and Tipper2009). Men may attend more to non-emotional aspects of faces – such as attractiveness – while women may attend more to emotional aspects. Men's empathy also appears to be mediated by social-cognitive factors that do not influence women's empathy. For example, similar brain regions are active when experiencing pain and when viewing pain in another individual for both sexes (Singer Reference Singer2006); however, men no longer display the same neural pattern when the other person is perceived as behaving unfairly (Singer Reference Singer2006). This suggests that men, compared to women, may more readily inhibit embodied emotion recognition strategies in such circumstances.

One encounters difficulties when trying to interpret the literature on sex differences in emotion recognition. Studies of mimicry often: (a) use only female participants, (b) use samples that are too small to allow comparisons of sex differences, (c) fail to test whether there are gender differences, or (d) fail to report the proportion of males and females in the sample. If there are only female participants, it remains untested whether effects would likewise occur in males. For example, females' embodiment during facial expression viewing has not been tested in male participants (Effron et al. Reference Effron, Niedenthal, Gil and Droit-Volet2006). Although others have assumed no gender differences, the body of work reported here suggests this assumption may be incorrect.

It would be fruitful to know whether encouraging men to mimic – rather than relying on their natural tendency to do so – improves their emotion recognition in either speed or accuracy. Interestingly, administering testosterone reduces facial mimicry in females (Hermans et al. Reference Hermans, Putman and van Honk2006); however, we know of no work which has examined whether increased testosterone impairs emotion recognition in either sex. This also raises the question of whether men's emotion recognition would be improved if they were instructed to attend to the eyes – something that women do naturally – since the frequencies of looks to the eyes is positively correlated with emotion recognition accuracy (Hall et al. Reference Hall, Hutton and Morgan2010). Developmental work is crucial for identifying the emergence of sex differences in face viewing patterns, mimicry, and emotion recognition, to determine the extent to which such behaviors and skills are continuous or changing throughout life.

We agree with Niendenthal et al. that embodied simulation is an important strategy for emotion recognition. We also concur that it is wise to examine each facial expression of emotion separately when modeling the processes underlying emotion recognition. In addition, we believe it is of utmost importance to consider individual differences – and especially sex differences – in the use of simulatory processes for each type of expression.