True and false smiles

Niedenthal et al.'s model is focused on the detection of true smiles of enjoyment, dominance, or affiliation. We are glad to see that Niedenthal et al. have acknowledged that the Duchenne smile (Ekman et al. Reference Ekman, Davidson and Friesen1990) is not the marker of a true smile. Duchenne smiles can be effortlessly posed, do not predict senders' affective state (Krumhuber & Manstead Reference Krumhuber and Manstead2009), and can be caused by negative experiences (e.g., losing a game: Schneider & Josephs Reference Schneider and Josephs1991; pain: Kunz et al. Reference Kunz, Prkachin and Lautenbacher2009). Nevertheless, Niedenthal et al. maintain the original theoretical assumption behind the Duchenne smile, namely, the true versus false dichotomy. According to Niedenthal et al., true smiles are honest displays that reveal positive feelings, whereas false smiles are manipulative displays that hide negative feelings.

In our view, any model aimed at detecting the “truth” or “honesty” of an expression is, ultimately, scientifically unsound. Smiles are not essentially true or false. Indeed, we doubt that any dichotomous classification of smiles will survive scientific scrutiny for long. Smiles are too morphologically diverse for a dichotomy: they can be located on a continuum based on their tempo, intensity, asymmetry, and concomitant facial movements (Krumhuber & Kappas Reference Krumhuber and Kappas2005; Krumhuber & Manstead Reference Krumhuber and Manstead2009). This dynamic diversity has consequences for the receiver's judgments (e.g., perception of trustworthiness: Krumhuber et al. Reference Krumhuber, Manstead, Cosker, Marshall, Rosin and Kappas2007) and is connected to the sender's relational strategies (e.g., Vigil Reference Vigil2009). Thus, each particular facial expression is too complex in its meaning to be either true or false. A consciously exaggerated “false” smile is “true” – and extremely functional – in the context of baby talk (Chong et al. Reference Chong, Werker, Russell and Carroll2003). The equally functional spontaneous “true” smile of a two-month-old baby is clearly “false” in that it lacks any of the “honest” emotional or cognitive correlates of a “true” adult smile.

Simulation as emotional experience

Niedenthal et al.'s simulation-as emotional-experience theory sees emotions as “natural kinds” in which experience and expression belong to a tight cluster of properties caused by the same neural structures (Barrett Reference Barrett2006; Fernández-Dols & Russell Reference Fernández-Dols, Russell, Millon and Lerner2003). On this view, emotion knowledge is expression (mimicry), experience, and recognition of basic emotion. There are two apparent sources of evidence in support of this view: experimental and clinical.

We question the experimental support (e.g., Halberstadt et al. Reference Halberstadt, Winkielman, Niedenthal and Dalle2009; Niedenthal et al. Reference Niedenthal, Winkielman, Mondillon and Vermeulen2009). Electromyography shows a weak or even contradictory relation between emotion knowledge and facial mimicry (e.g., activity of the levator labii when processing joy or nonspecific activity of the corrugator when processing disgust; Niedenthal et al. Reference Niedenthal, Winkielman, Mondillon and Vermeulen2009). The claim that such mimicry causes the experience of the corresponding basic emotion is not scientifically justified on the basis of current evidence. We see no reason to assume that the link between emotion knowledge and facial mimicry is different from the observed link between nodding your head and emotion knowledge (Niedenthal Reference Niedenthal2007) – that is, culture-specific and learned. No evidence forces us to the additional assumption of an innate connection between mimicry, experience, and recognition of emotion. Mimicry could be the outcome of several parallel processes (Lishner et al. Reference Lishner, Cooter and Zald2008; Yabar et al. Reference Yabar, Johnston, Miles and Peace2006).

We also question the clinical support for the recognition-as-experience view. Case studies show paired deficits in both experience and recognition of emotions (e.g., fear or disgust) apparently caused by damage in one specific brain region (e.g., the amygdala or the insula and basal ganglia; for a review, see Goldman Reference Goldman2006). Nevertheless, further research on the shared neural substrate of such paired deficits has raised doubts as to whether these regions are organized for experiencing and recognizing basic emotions. The functions of these regions are broader (e.g., Heberlein et al. Reference Heberlein, Padon, Gillihan, Farah and Fellows2008), more diverse (e.g., Heberlein & Atkinson Reference Heberlein and Atkinson2009), and more complex than originally thought (e.g., Chen et al. Reference Chen, Dammers, Boers, Leiberg, Edgar, Roberts and Mathiak2009; Straube et al. Reference Straube, Weisbrod, Schmidt, Raschdorf, Preul, Mentzel and Miltner2010). Better experimental procedures and measures challenge the original interpretation of the paired deficits (Milders et al. Reference Milders, Crawford, Lamb and Simpson2003; Rapcsak et al. Reference Rapcsak, Galper, Comer, Reminger, Nielsen, Kaszniak, Verfaellie, Laguna, Labiner and Cohen2000). Understanding recognition requires a more conceptually sound analysis of the dynamics and functions of specific brain regions (see Kagan Reference Kagan2007).