Henrich et al. present a compelling empirical case for a reduction in the current reliance on participants from WEIRD societies in the study of the human mind. Awareness of these facts should motivate investigators both to recognize the need for caution in inferring the generalizability of results derived from parochial samples and to pursue more diverse samples, whether directly or through collaborative networks. Here I wish to suggest that additional impetus for such changes derives not simply from questions of representativeness, but rather from the fact that investigators themselves inevitably view the world through the lens of their own culture.
As Henrich et al. observe, most psychological research is conducted by members of WEIRD societies, hence investigators are largely studying people very similar to themselves. By virtue of the fact that they, too, are enculturated humans, investigators will often share folk psychological models with participants drawn from their own cultural group. Although rigorous science is always based on the refinement of models through hypothesis testing, one has to begin somewhere, and I strongly suspect that social scientists in general, and psychologists in particular, often rely on their own folk models as a starting point in this process. When investigators share fundamental cultural commonalities with their participants, they run the risk of overlooking key features of the psychological phenomena at issue, as such features may be absent from, or downplayed by, the given folk models (see Levy Reference Levy1973).
In light of the above, it is interesting to consider a domain of research only addressed in passing by Henrich et al., namely, the study of emotions. Due, in part, to the centrality of claims regarding innateness in a number of seminal modern investigations of emotion, building on Darwin's (Reference Darwin1872) prior work on the subject, the study of emotion has long included a substantial cross-cultural component. From early on, cross-cultural research was employed to explore both the recognizability of emotional expressions (e.g., Ekman & Friesen Reference Ekman and Friesen1971) and the qualia and elicitors associated with different emotions (e.g., Wallbott & Scherer Reference Wallbott and Scherer1986).
Despite this emphasis, the relative neglect by prominent universalists of the importance of cultural meaning systems in the experience of emotion eventually led to bold claims by some cultural constructionists (e.g., Lutz Reference Lutz1988), wherein a panhuman psychology was seen as merely the seed from which radically diverse parochial forms of emotion spring. As the field matured further, polar differences between nativists and constructivists eventually dulled, with various middle grounds being found, wherein investigators sought to systematically examine the nature and extent of cultural variation in the elicitors of emotions, the non-emotion correlates of cultural variation in emotional experience, and similar facets (see, e.g., Kitayama & Markus Reference Kitayama and Markus1994; Mesquita & Frijda Reference Mesquita and Frijda1992).
While the psychology of emotion would thus seem to be one area where the call to action trumpeted by Henrich et al. has already long been heard, even here one can find signs of an underestimation of the importance of moving beyond cultural similarities between investigator and participant. For example, a burgeoning literature (a keyword search in PsycINFO produces almost 6,000 hits) explores shame. This literature focuses on self-consciousness and issues of moral and personal worth, often contrasting shame with guilt (see Tangney & Dearing Reference Tangney and Dearing2002). However, results obtained from a small-scale non-WEIRD society, and hints provided by many non-Western languages' emotion lexicons, suggest that, in addition to the aforementioned features, for many of the world's peoples, the emotion of shame also encompasses what English speakers would call respect and fear, facets that primarily concern subordinance in a hierarchy rather than failure to conform to social standards (Fessler Reference Fessler2004).
Although the subordinance aspect of shame is absent from the vast majority of the voluminous scholarly work on the subject, with guidance, English-speaking participants generally recognize this facet of shame in their own experiences. Importantly, however, they do not volunteer this association on their own. This is not surprising, given that Western folk models of shame entirely ignore the experience of subordinance, perhaps because this experience is arguably antithetical to the values common in WEIRD groups. This leads to my central question, namely, why have investigators not attended more to the subordinance facet of shame?
I suspect that the problem is not merely that the participants studied by many students of shame differ in important ways from most of the world's peoples, but also that the investigators studying those participants consistently find strong support for their own intuitions, and hence see less need to cast a wide net despite the history of cross-cultural research on emotions. Whereas early work on emotions included places such as New Guinea villages (Ekman & Friesen Reference Ekman and Friesen1971), and recent work on pride, directly relevant to understanding shame, has been conducted in rural Burkina Faso (Tracy & Robins Reference Tracy and Robins2008), much current work on shame is confined to educated and affluent members of the state-level societies of North America, Western Europe, and, to a lesser extent, East Asia. Importantly, failure to recognize those aspects of shame that are downplayed or ignored in their own culture limits investigators' ability to explain key features of the phenomena of interest. For example, the behavioral tendencies for flight and hiding so prominent in the experience of shame seem odd in the context of a moral emotion, but are readily understood once it is recognized that this moral emotion is built upon a simpler emotion active in the dangerous context of dominance negotiations (see Fessler Reference Fessler, Tracy, Robins and Tangney2007).
The take-home lesson here is that we must be particularly cautious to avoid employing participants who are culturally similar to ourselves whenever the given enterprise involves testing predictions that resonate with our intuitions. Such samples are unlikely to reveal to us that which we do not know that we do not know, features that are masked by our own folk models.
You have
Access
Henrich et al. present a compelling empirical case for a reduction in the current reliance on participants from WEIRD societies in the study of the human mind. Awareness of these facts should motivate investigators both to recognize the need for caution in inferring the generalizability of results derived from parochial samples and to pursue more diverse samples, whether directly or through collaborative networks. Here I wish to suggest that additional impetus for such changes derives not simply from questions of representativeness, but rather from the fact that investigators themselves inevitably view the world through the lens of their own culture.
As Henrich et al. observe, most psychological research is conducted by members of WEIRD societies, hence investigators are largely studying people very similar to themselves. By virtue of the fact that they, too, are enculturated humans, investigators will often share folk psychological models with participants drawn from their own cultural group. Although rigorous science is always based on the refinement of models through hypothesis testing, one has to begin somewhere, and I strongly suspect that social scientists in general, and psychologists in particular, often rely on their own folk models as a starting point in this process. When investigators share fundamental cultural commonalities with their participants, they run the risk of overlooking key features of the psychological phenomena at issue, as such features may be absent from, or downplayed by, the given folk models (see Levy Reference Levy1973).
In light of the above, it is interesting to consider a domain of research only addressed in passing by Henrich et al., namely, the study of emotions. Due, in part, to the centrality of claims regarding innateness in a number of seminal modern investigations of emotion, building on Darwin's (Reference Darwin1872) prior work on the subject, the study of emotion has long included a substantial cross-cultural component. From early on, cross-cultural research was employed to explore both the recognizability of emotional expressions (e.g., Ekman & Friesen Reference Ekman and Friesen1971) and the qualia and elicitors associated with different emotions (e.g., Wallbott & Scherer Reference Wallbott and Scherer1986).
Despite this emphasis, the relative neglect by prominent universalists of the importance of cultural meaning systems in the experience of emotion eventually led to bold claims by some cultural constructionists (e.g., Lutz Reference Lutz1988), wherein a panhuman psychology was seen as merely the seed from which radically diverse parochial forms of emotion spring. As the field matured further, polar differences between nativists and constructivists eventually dulled, with various middle grounds being found, wherein investigators sought to systematically examine the nature and extent of cultural variation in the elicitors of emotions, the non-emotion correlates of cultural variation in emotional experience, and similar facets (see, e.g., Kitayama & Markus Reference Kitayama and Markus1994; Mesquita & Frijda Reference Mesquita and Frijda1992).
While the psychology of emotion would thus seem to be one area where the call to action trumpeted by Henrich et al. has already long been heard, even here one can find signs of an underestimation of the importance of moving beyond cultural similarities between investigator and participant. For example, a burgeoning literature (a keyword search in PsycINFO produces almost 6,000 hits) explores shame. This literature focuses on self-consciousness and issues of moral and personal worth, often contrasting shame with guilt (see Tangney & Dearing Reference Tangney and Dearing2002). However, results obtained from a small-scale non-WEIRD society, and hints provided by many non-Western languages' emotion lexicons, suggest that, in addition to the aforementioned features, for many of the world's peoples, the emotion of shame also encompasses what English speakers would call respect and fear, facets that primarily concern subordinance in a hierarchy rather than failure to conform to social standards (Fessler Reference Fessler2004).
Although the subordinance aspect of shame is absent from the vast majority of the voluminous scholarly work on the subject, with guidance, English-speaking participants generally recognize this facet of shame in their own experiences. Importantly, however, they do not volunteer this association on their own. This is not surprising, given that Western folk models of shame entirely ignore the experience of subordinance, perhaps because this experience is arguably antithetical to the values common in WEIRD groups. This leads to my central question, namely, why have investigators not attended more to the subordinance facet of shame?
I suspect that the problem is not merely that the participants studied by many students of shame differ in important ways from most of the world's peoples, but also that the investigators studying those participants consistently find strong support for their own intuitions, and hence see less need to cast a wide net despite the history of cross-cultural research on emotions. Whereas early work on emotions included places such as New Guinea villages (Ekman & Friesen Reference Ekman and Friesen1971), and recent work on pride, directly relevant to understanding shame, has been conducted in rural Burkina Faso (Tracy & Robins Reference Tracy and Robins2008), much current work on shame is confined to educated and affluent members of the state-level societies of North America, Western Europe, and, to a lesser extent, East Asia. Importantly, failure to recognize those aspects of shame that are downplayed or ignored in their own culture limits investigators' ability to explain key features of the phenomena of interest. For example, the behavioral tendencies for flight and hiding so prominent in the experience of shame seem odd in the context of a moral emotion, but are readily understood once it is recognized that this moral emotion is built upon a simpler emotion active in the dangerous context of dominance negotiations (see Fessler Reference Fessler, Tracy, Robins and Tangney2007).
The take-home lesson here is that we must be particularly cautious to avoid employing participants who are culturally similar to ourselves whenever the given enterprise involves testing predictions that resonate with our intuitions. Such samples are unlikely to reveal to us that which we do not know that we do not know, features that are masked by our own folk models.