limited the scope of these findings (Ekman et al., 1987).
In a second type of study, American and Japanese men were secretly filmed while watching either an emotionally neutral nature video or one depicting bodily mutilation (Ekman, 1972). The men viewed the films either alone or with an experimenter present. Researchers predicted that Japanese men would show a disgust reaction to the mutilation film when alone but would suppress it when another person was present. In contrast, the Americans were expected to express their emotions whether or not they were alone. The findings supported the hypothesis and demonstrated that the two cultures have different display rules dictating when it is appropriate to show certain facial expressions (C. M. Moran, Diefendorff, & Greguras, 2013). The Japanese participants clearly felt disgust, as shown in the “alone” condition, but chose not to show this when an observer was present. The cultural boundedness of the display rule was demonstrated by the fact that the Americans made no such attempt to hide their disgust. This study was important because it was the first to empirically show differences in display rules; however, its generalizability is limited because only two cultures and two expressions were investigated (Ekman et al., 1987).
In the third and most important type of study, Ekman and other researchers exposed individuals from 12 cultures to photographs of facial expressions (Ekman, Sorenson, & Friesen, 1969; Izard, 1971) and found high levels of agreement across the six major emotions. In a follow-up study, Ekman found agreement in 10 cultures—including the United States, Japan, Sumatra, Estonia, and Turkey—on the same six primary emotions.
More recently Elfenbein and Ambady (2002) conducted a meta-analysis of 97 studies encompassing 42 cultures and over 20,000 participants. A meta-analysis combines multiple studies—usually conducted by different researchers—into one analysis that allows the researcher to look at larger sets of data than are included in any single study. The researcher can have greater confidence in the validity of conclusions drawn from this analysis of analyses that considers all of the data than from just one of these studies. Elfenbein and Ambady’s findings validated earlier research with regard to the six core emotions—anger, surprise, fear, happiness, sadness, and disgust. Interestingly, they also found that individuals were generally better at correctly recognizing expressions from individuals who were in their same region, ethnic, or national group versus those in different groups (Elfenbein & Ambady, 2002). Overall, while there is some universality in the expression of emotions, much cultural variability also exists (Haidt & Keltner, 1999; Izard, 1994; Perrett, 2012).
Cross-cultural similarities in emotional expressions, including groups with no previous contact with the outside world, provide strong evidence for Darwin’s claim about the innateness of emotional expressions. Other research has extended the findings described above. For instance, Tracy and Robins (2007) found many commonalities among pride displays in the United States, Italy, and in an isolated, preliterate West African tribe. Pride is associated with meaningful achievements and is often displayed with a low-intensity smile, head tilt, expanded posture, and hands either raised above the head or resting on the hips (Tracy & Robins, 2007; Tracy, Shariff, Zhao, & Henrich, 2013). Furthermore, one analysis revealed that sighted, blind, and congenitally blind athletes from over 30 countries demonstrated many similarities in their pride expressions following Olympic or Paralympic success (Matsumoto & Willingham, 2009; Tracy & Matsumoto, 2008). Tracy and Matsumoto (2008) also found that shame—in this case resulting from athletic failure—was displayed in similar ways among all of the groups, although the displays were moderated (weaker) in persons from more individualistic cultures.
Finally, Keltner and Buswell (1997) noted a number of cross-species commonalities in expressions of embarrassment or what they call appeasement behavior, such as gazing down, covering the face, and contraction or shrinkage of posture (Keltner, 1995). According to these theorists, appeasement follows the violation of a social norm and is essentially a display of submissive and affiliative behaviors designed to solidify social relations. In sum, the fact that the primary emotions are similar across cultures supports evolutionary theory, but the existence of some cross-cultural variability reinforces the notion that evolution is not the whole story.
One potentially significant qualification is in order here. Although the cultural universality of the six basic emotions has been widely recognized and generally assumed, very recent research has called this into question. Several researchers have argued that the few studies that have included truly remote cultures—previously without contact with the West—have a flawed methodology (Jack, 2013; Jack, Blais, Scheepers, Schyns, & Caldara, 2009). For instance, studies by Ekman and others typically supplied emotion words along with the facial expressions and asked participants to match them up. It is possible that, by providing the emotion words, the researchers shaped the responses of non-Western participants in a way that made them appear to be consistent with the Western participants. However, recent studies of remote, non-Western cultures without previous contact with the West in which emotion words were withheld found important differences in emotion recognition (Barrett, Mesquita, & Gendron, 2011; Gendron, Roberson, van der Vyver, & Barrett, 2014). If this research is replicated and holds up to scrutiny, then psychologists will have to rethink this claim to universality. The fact that psychologists continue to examine and reexamine the assumptions and findings of prior research reflects one of my favorite things about science: its self-corrective nature.
Nonverbal Behavior: Perceptible social behavior that is extra linguistic and not primarily intended to manipulate the physical world
Display Rules: Rules indicating which facial expressions are appropriate in a given context
Meta-analysis: Combines multiple studies, usually by different researchers, into one analysis that allows the researcher to draw conclusions about the set of studies as a whole
Recognizing Happiness
Would you predict that humans are faster at identifying a happy face or an angry face when the face is embedded in a crowd of non-happy or non-angry faces? Recall that Darwin argued that the ability to both read and signal emotions via facial expressions provide an adaptive advantage. It would seem, then, that individuals who quickly read and appropriately respond to facial expressions could more successfully navigate the social landscape. Therefore, you might expect that we would be adept at noticing expressions suggestive of potential threats to our well-being, such as angry faces (Reed, DeScioli, & Pinker, 2014). Several years ago, when I began writing this textbook, psychological research on facial recognition supported this notion. For instance, Hansen and Hansen (1988) asked participants to identify, as quickly as possible, a single divergent expression from a sea of faces. In some cases a happy face was hidden in a sea of neutral or angry faces, whereas in others an angry face was placed with neutral or happy faces. Almost invariably, angry faces were identified much more rapidly than happy or neutral faces. Hansen and Hansen (1988) dubbed this the face-in-the-crowd effect and found it whether the crowd faces were all of the same or different persons (Schmidt-Daffy, 2011).
However, relatively recent research has not only called this effect into question but has demonstrated the opposite one: People are actually faster at finding happy faces! Becker and Wright (2011) ran seven studies that reversed the way psychologists view facial recognition. Since the details about how they did this are too complicated to delve into here, I will give you the short version. Becker and Wright identified and eliminated a number of confounds with the earlier face-in-the-crowd studies that led the earlier researchers to mistakenly believe that the differences in their dependent variables were caused by the manipulations of their independent variables. Once Becker and Wright altered the experimental procedures, they found that smiling, happy faces were easier to spot. This was true even if they used closed-mouth smiles in which teeth were hidden. These researchers hypothesize that people are better at detecting happy faces because they are far more prevalent than are angry expressions and that happy faces have evolved to be more easily recognizable.
Emotion Blends and Dialects
Given that facial expressions and other nonverbal behavior play a key role in the development and maintenance of social relations (Ambady & Weisbuch, 2010; Keltner & Kring,