Culture as situated cognition: Cultural mindsets, cultural fluency, and meaning making

Abstract

Culture is a human universal, a “good enough” solution to universal needs. It is also a specific meaning-making framework, a “mindset” that influences what feels fluent, what is attended to, which goals or mental procedures are salient. Cross-national comparisons demonstrate both universality and between-group difference (specificity) but cannot address underlying process or distinguish fixed from context-dependent effects. I use a situated cognition framework and experimental methods to address these gaps, demonstrating that salient cultural mindsets have causal downstream consequences for meaning making, self-processes, willingness to invest in relationships, and complex mental procedures. Moreover, individualistic and collectivistic mindsets are accessible cross-culturally so both can be primed. Between-group differences arise in part from momentary cues that make either individualistic or collectivistic mindset accessible.

Keywords:

• Culture
• Situated cognition
• Individualism
• Collectivism

Notes

¹ I use the term cultural mindset to mean a set of mental representations or cognitive schema containing culture-congruent mental content (knowledge about the self and the world), cognitive procedures (e.g., “find relationships and connect” or “find main point and separate”) and goals (e.g., “fit in and be sensitive to context” or “stick out and do your own thing”) (Oyserman & Lee 2008a, 2008b; Oyserman, Sorensen, Reber, & Chen 2009). When in an individualistic mindset people attend to content, procedures, and goals relevant to distinction; when they are in a collectivistic mindset people attend to content, procedures, and goals relevant to connection

² A number of research projects have provided country-level estimates of individualism and collectivism scores. Estimates are typically based on convenience samples of participants and use their responses to value scales (Gelfand et al., 2004; Hofstede, 1980, 2001; Suh, Diener, Oishi, & Triandis, 1998), although one scale focuses on language structure (Kashima & Kashima, 1998). There are also limitations to this approach including slippage between the abstract constructs and specific operationalisation, over-reliance on explicit self-report, concerns about social desirability, and response style differences (for a review of limitations, see Oyserman, Coon, et al., 2002). Moreover, individualism and collectivism values are not fixed, but change with societal wealth so that the assumption that data obtained at an earlier time fit current data or that contemporary data fit an earlier time is not strictly valid (Hofstede, 1980, 2001; Inglehart & Oyserman, 2004; Kağıtçıbaşı, 1997). Of course societies that differ in individualism and/or collectivism values do not necessarily differ in other values (Fischer & Schwartz, 2010).

³ With regard to bribery, Mazar and Aggarwal (2011) show both cross-national correlational evidence and mindset priming evidence supporting the association between collectivism and bribery (of out-group members). Primed collectivism reduces sense of personal responsibility, which mediates the effect of collectivism on willingness to bribe. With regard to pathogens, Fincher and colleagues (2008) show that historical and current pathogen levels are higher in regions that are currently higher in collectivism, arguing that when risk of infection and disease is higher people should feel more threatened by out-group members and feel more dependent on in-group members. These data use secondary sources and are correlational so causal argument is not possible.

⁴ Effect sizes are reported following the recommendations of J. Cohen (1988) in interpreting the meaning of the observed effect sizes, effect sizes of less than d = 0.2 are described as “small”, those of d = 0.5–.7 are described as “moderate” and those above d = 0.8 as “large”.

⁵ Puente-Diaz (2011) provides an interesting conceptual replication, with Mexican adults in Mexico City. Adults who were asked about their life satisfaction and then about their satisfaction with their romantic life gave answers correlated at r = .50; the correlation dropped significantly to r = .36 when the specific question was asked first. Thus results for Mexicans parallel those for Chinese, implicating sensitivity to the conversational norm among Mexican participants.

Additional information

Notes on contributors

Daphna Oyserman

With thanks for funding support from the Center for Advanced Study in the Behavioral Sciences, Stanford, California, and the Alexander von Humboldt Foundation, Germany.