Theoretical and empirical content
The first two articles discuss the conceptions and influence of culture on social behavior. Paez & Gonzalez’s article discusses the way culture can be conceived, and explains differences in social behavior. Ovejero’s article proposes a social constructionist view of emotions. The first article stresses a realistic and sociocultural approach to culture, cognition, motivation and emotion. The second posits the importance of language and historical and cultural relativism in the understanding of emotions.
Gouveia and Ros present Hofstede and Schwartz’s conceptions concerning cultural dimensions of values, and show the limitations of conceiving individualism-collectivism as the most important dimension and as one with two poles. Differences between individualistic developed countries (such as the USA and Western Europe) are important, and suggest the relevance of cultural dimensions other than individualism-collectivism. Morales et al.’s article highlights the multidimensionality of individualism and presents data showing that Spanish respondents share other meanings of individualism, different from the self versus ingroup conflict (an emic Western idea believed to be true for all world culture).
Finally, Garzón analyzes the issue of familism. This is a concept used for analyzing the change towards family values, and is a central tenet in cultural change. Garzón presents empirical data from current Spanish society stressing the impact of familism in a series of cultural dimensions of belief systems, from the point of view of historical development, knowledge and society. Familism implies lower civic culture and amoral familism, a mixture of rugged individualism beyond the familial group, nepotism and localism. Garzón shows empirical data that operationalize these cultural beliefs and shows its importance in contemporary spanish society.
In the third section, five articles examine how culture influences emotional experience and regulation. Basabe et al. examine the influence of ecological (climate), socioeconomic development and cultural dimensions on the self-reported experience of emotion (frequency, intensity and social desirability), using meta-analysis and integrating five large cross-cultural studies. Ubillos et al. also examine the sociocultural determinants of frequency of coital behavior and extramarital affairs, as indices of sexual activity and sexual permissiveness. Fernández et al. use a similar approach to explain differences between countries in self-reported emotional verbal and nonverbal reactions. Singh examines gender and cultural variations in the social sharing of emotions, comparing Indian, migrant Indian and British adolescent samples. Finally, Etxebarría exhaustively reviews the literature on culture and one of the most important emotions for social regulation: guilt.
The four empirical studies confirm important hypotheses and challenge some assumptions. The studies partially support classic Marxist explanations: Basabe et al. found that high socioeconomic development was related to well-being, introversion and intensity of emotions, and Fernández et al. confirm that economic development is related to greater emotional expression. In the same vein, Ubillos et al. show that strong economic development is related to higher frequency of sexual behavior. This suggests that increases in quality of life, privacy and social resources related to levels of earnings, education and life expectancy, lead to a more positive and intense emotional experience.
The studies also confirm classic Weberian and Durkheimian hypotheses: culture determines emotional experience. Hofstede’s (1991) cultural dimensions were associated, in a congruent way, with emotional feeling, expression and sexual behavior. Societies high in power distance, which legitimize differences in social status and stress «respect», present a negative and repressive emotional culture: lower subjective well-being, less sexual permissiveness (i.e. lower frequency of extramarital sex for both men and women), high frequency of negative emotions, high social rejection and lower intensity of both negative and positive emotions and, congruently, lower verbal and non-verbal expression of emotions. Challenging the centrality of the individualism-collectivism dimension, Basabe et al.’s article concludes that power distance is one of the most important correlates of emotional life.
Studies also show an interesting interaction between gender differences and the power distance cultural dimension. Fernández et al. found that women were always more expressive than men. Singh also confirms, in general, higher levels of social sharing of emotions in females than in males. However, Asian participants report the lowest level of gender differences in emotional expression. This suggests a strong normative system of emotional display rules in Asian cultures, characterized by high power distance. Ubillos et al. found similar results for sexual permissiveness. In India, a society that also has a high power distance culture, male adolescents report more intense emotional reactions and social sharing than females. Similar results were found by Pandey et al. (1996): in India females score higher than males in the alexythimia sub-scale «difficulty in describing feelings». This pattern of higher emotionality and emotional verbal expression in Indian males is congruent with another study showing that friendship between Indian males proved to be as intense as that between Indian females, while in the United States female same-sex friendships were more intense than males ones (Berman, Murphy-Berman and Pachauri, 1988, quoted in Moghaddam, 1998, p.261). Using data from 10 countries, Páez et al. (1999) found that cultural dimensions explain the variability in gender differences in relation to verbal expression of emotions. In this sense, in high power distance and collectivist cultures women score higher than men in difficulty for describing feelings and the opposite occurs in individualistic countries. This result is probably explained by norms inducing lower female participation in public life, including a lower level of public vocalization. In any case, as Singh argues, it is not possible to explain gender differences as a mere assimilation of women to collectivism and men to individualism.
Different articles also confirm the importance of another dimension beyond individualism-collectivism, the Femininity-Masculinity cultural dimension (Hofstede, 1998). Societies high in cultural Masculinity, that stress competition, show a negative emotional climate: lower well-being and higher frequency of negative emotions (Basabe et al.), lower frequency of sexual behavior (when controlling for sociostructural factors, Ubillos et al.), or lower emotional expression (Fernández et al.). However, cultural Masculinity was not related to normative inhibition or to the intensity of emotional feeling (Basabe et al.).
High uncertainty-avoidance societies, which stress regulation of ambiguity, show low subjective well-being, stress emotional normativeness and were associated with high intensity of emotions and sexual behavior, even if these results were valid only in Europe (not in America). In the study on sexual behavior most of the high uncertainty-avoidance countries sampled were European.
Individualism was associated with well-being, positive and negative emotional intensity and social desirability, higher frequency of sexual behavior, and higher emotional expression, but not with frequency of positive emotions. Fernández et al. also show important differences between Asian and Latin-American collectivist subjects. Asian collectivists report generally lower emotional expression than Europeans and North Americans (USA). Latin-Americans report, in a similar way to Asians, lower expression of negative emotions but higher levels of expression of positive emotions, thus coinciding with European and USA samples. Etxebarría examines the role of guilt on self-regulation and social control. Collectivist cultures are assumed to be shame cultures and individualistic ones guilt cultures. Her data partially challenge these assumptions: both guilt and shame are less frequent and have fewer effects on self-regulation (i.e., lower negative influences on self-esteem and relationships) in collectivist and high power distance cultures. In individualistic societies shame more typically resembles guilt, and this result confirms the importance of self-regulation in these cultures. Etxebarría discusses both this and other possible effects of culture on guilt.
The fourth main topic is cultural differences in coping with traumatic events. Two studies analyzed intra-country cultural differences: Mapuches versus Mestizos in Chile, and Mayas versus Mestizo Ladinos in Guatemala. Pérez et al.’s article presents a cross-cultural study on bereavement and long-term adaptation to traumatic events, and Martín et al. examine the role of rituals, coping by means of social sharing and silence, in the case of the Guatemalan genocide. Relatively more collectivist indigenous cultures were associated with higher resilience when confronting negative social situations: lower community crisis (Martín et al.) and lower affective disorders related to traumatic events (Pérez et al.). Martín et al. found that subjects belonging to traditional Mayan culture reported lower anger-related reactions and coping reactions than Spanish-speaking Guatemalan subjects. Pérez et al. also found that traditional Mapuches show a lower level of affective disorders than more acculturated Mapuches. These results confirm that collectivist and high power distance cultures such as the Mayan one regulate more negative emotional reactions, a finding congruent with the third section of studies showing that negative emotional intensity and expression were lower in high power distance and collectivist countries. Participation in funeral rituals also showed stronger buffer effects in the case of Mayan subjects (participation decreased anger/sense of injustice and did not reinforce the quest and demand for commemorations among the Mayas). The article discusses different explanations and implications of these results. Martín et al.’s results disconfirm that silence and emotional inhibition were more typical of collectivist Mayas, but reaffirm that Mayas report lower levels of coping by means of talking or social sharing. Interpersonal communication reinforces anger responses and demands for commemorations (less frequent responses among Mayas), suggesting that social sharing is more important in collectivist cultures, where emotional verbalization is less frequent than in more individualistic cultures, which value verbal expression of emotion. This articles also examines factors predicting intense grief, and points out implications for human rights and mental health interventions, showing how basic and applied social psychological research could, and should, be articulated.
Contributions also pose and examine methodological problems encountered in cross-cultural research. Some of these issues are as follows:
a) The assumption that the samples used represent the culture of a nation and that the sampling of nations includes cultures that are really different. Some studies (e.g., Diener et al., 1995 and Pennebaker et al., 1996) used non-representative samples, and this is a methodological weakness. However, student sample means of affect (subjective well-being) correlated well with means based on representative samples (Diener, Diener & Diener, 1995). In Basabe et al.’s study, means based on student samples show a satisfactory concurrent validity with other studies using representative samples. These results confirm the validity of the nation’s affect means in studies using smaller-sized convenience samples. As Ovejero argues, psychology and social sciences students share a middle-class and scientific subculture around the world: they are relatively homogenous and not representative of the general population of each country. On the other hand, student’s non-representative samples may present the position of the cultural group relative to one which is similar (e.g., another student sample found in other groups). Showing differences in these samples sharing social class and professional subculture is a strength rather than a weakness of current cross-cultural research. Finally, a wide series of cross-cultural research used representative samples.
b) Construct validity and the semantic similarity or meaning equivalence of stimuli or instruments across countries and time (see Ubillos et al. regarding sexual behavior as coital behavior, for instance). The question of semantic construct validity across cultures is a common source of criticism. One important assumption is that the means of individuals’ responses can be used to compare nations because meanings of values and behaviors are relatively similar (this ignores the possibility that different meanings may be associated with abstract values and concrete behaviors). Some authors question that cultures can be compared along a common dimension, because there are no common value dimensions (power distance may be a salient dimension in one group of countries but irrelevant in another group) (McCauley, Ottati & Lee, 1999). Ethnographical accounts and theoretical analysis could be used to demonstrate the relevance of the dimension. Hofstede’s dimensions (1991) are solutions to general problems of social coordination and, probably, hierarchical and power arrangements represent a general problem relevant to all societies. There are also a variety of empirical procedures for testing dimensional relevance, using associations between items (meaning is represented by the association between statements, adjectives, values, etc.), similar and satisfactory reliability coefficients, similar and congruent results in factor analysis (confirmatory in particular), cluster analysis and other multidimensional procedures. These methods all serve to check the dimensional relevance of instruments. Schwartz’s study on values (1994) shows strong meaning construct validity by means of multidimensional analysis. As a response to meaning equivalence it is important to notice that results reported in the articles were usually triangulated by means of an independent series of studies, and the general pattern shows convergence. Criticism related to the historical evolution of countries is another important issue. Basabe et al. present some data suggesting that some countries have increased their level of individualism. However, the trend towards high individualism is a general one related to industrialization and modernization. Hofstede’s data (1991) was collected during the formative years of current young and adult samples. Theoretical approaches, the generational hypothesis, Inglehart’s socialization hypothesis (1991) and empirical data confirm that cultural knowledge is acquired essentially during childhood (Sangster & Reynolds, 1996) and in adolescence and young adulthood (Pennebaker, Páez & Rimé, 1997).
c) Construct validity of instruments and content validity: the assumption that instruments tap most of the important values of all cultures or emotional features and that countries selected are representative of the world’s most important cultures. Another related aspect is the problem of historical evolution (see Ovejero’s article). Hofstede’s «values» items (1991) were heterogeneous in content; some items tapped work goals rather than values, and the interpretation of factors does not follow easily from the items content. However, Hofstede assembles ample evidence supporting the construct validity of the values dimensions (Peabody, 1999). Moreover, with respect to values items and content validity, Fiske, Markus, Kitayama and Nisbett (1998) conclude that four different sets of research on values (Hofstede, Schwartz, Troompenars and Chinese Culture Connection) converge on the individualism-collectivism and power distance cultural dimensions, including one set of values based on Chinese culture. These results suggest that Hofstede and Schwartz’s instruments overlap with Chinese indigenous values. Other authors criticize Hofstede’s scores because the positions of some countries appear as incongruent with stereotypes and ethnographical accounts of these countries. For instance, supposedly «macho» countries such as Spain, Portugal and some Latin-American countries score lower in masculinity (Peabody, 1999). Nevertheless, these countries, in spite of relatively strong gender differences, stress cooperation and social support and de-emphasize challenge (Costa Rica, Uruguay, Chile, Panama, Peru, Guatemala fit this description well). Other «macho» countries (Mexico, Colombia, Venezuela and Ecuador) had higher scores on masculinity. Basabe et al. and Ubillos et al. also deal with the problem of cross-cultural validity of affect measures and sexual behavior indices. Fernández et al.’s prototypical emotional features were based on large-scale European and American studies, and a content analysis of Asian emotional accounts was also performed to check how exhaustive and universal the items were. A more important criticism concerns the nation’s sample. Hofstede’s sample (1991) included a large group of American and Western European countries and some countries from Asia, with a notable lack of African and Eastern European ones. Schwartz’s sample (1994) includes a larger and more satisfactory group of countries. Basabe et al.’s meta-analysis shows strong content validity for countries sample. Other studies, such as those of Gouveia and Ros, Ubillos et al. and Fernández et al. analyze only some 20-25 countries and, for example, the association between «southernness» or hot climate, uncertainty avoidance and emotional expressiveness and more intense sexual activity appear as valid only for European countries, showing how a restriction in the countries’ samples affects general relationships. On the other hand, Gouveia and Ros found strong concurrent validity between Hofstede and Schwartz’s nations’ cultural values scores and strong construct validity with macrosocial indices. Basabe et al., Ubillos et al. and Fernández et al.’s studies in the section of culture and emotion also found strong correlations between Hofstede’s nation values scores and macropsychological indices (nations’ means on subjective well-being, reported emotional intensity, frequency and social desirability and sexual behavior) or psychological indices (individuals’ self-reports on emotional expression). In this way, Basabe et al. confirm that self-reports of affect balance had satisfactory concurrent validity. Globally, cultural values scores show convergent validity with current surveys of values and current macrosocial and macropsychological indices.
d) Another question is the assumption that means of individual responses reflect collective values, or the problem of using aggregated data or means of psychological self-report data as valid indices of collective processes, the associated consequence of not taking into account intra-nation variability and the problem of assimilation of nations to cultures (see Fernández et al.). Morales et al. posit the importance of intra-nation variability and present data showing high intra-nation variability in individualism with Spanish respondents. Pérez et al. and Martín et al. examine intra-nation cultural variability using self-reported traditionalism and mother tongue as a crude but operational cultural marker. One possibility is to check whether there is an intra-nation convergence of responses by means of the intraclass coefficient (Kenny & LaVoie, 1985). This contrast is not frequently used, even though it generally confirms the interdependence of nation samples for psychological variables. For instance, convergence of self-reported perceived emotional climate was tested by the intraclass coefficient, confirming the interdependence responses within groups, and this statistical test is evidence of the validity of aggregated data or group mean as a macropsychological indicator (Páez et al., 1997). In order to analyze data, Fernández et al., rather than using means across nation samples, used all subjects in the respective samples and collapsed nations in level of cultural dimensions, overcoming the identification of «one nation=one culture» problem and providing a strong test of significant effects, since within-country sample variance was taken into account. However, Basabe et al., in another study (1999), check the effects of cultural dimensions on individual responses and not on nations’ means (on self-reported affect balance in data from two large cross-cultural matrices from Diener’s et al. and Pennebaker et al.’s studies), and replicate, at an individual level, the same effects, suggesting that at least for reported emotional feeling and expression, collective and psychological processes are similar.
e) A further problem is the assumption that values are an important feature of cultures related in some ways to institutions and individual practices, or the problem of functional equivalence of stimuli. In other words, the problematic relationship between values and social behavior (which psychosociological research shows is usually low), or the real importance of declarative conscious knowledge for actual behavior. Culturalist explanations usually ignore the importance of situation and structural aspects. However, in all of the empirical studies, multivariate analyses were performed, usually in order to disentangle the influence of socioeconomic development and ecological factors from cultural dimensions. The influence of direct perception (frequency of emotional episodes) and social norms related to values (perceived social desirability of emotions) were used as mediating processes explaining the influence of cultural values and structural factors on emotions (see Basabe et al.). Other studies rely on macrosocial indicators (such as GDP) and self-reported aggregated mean measures of behavior (see Gouveia et al. and Ubillos et al.), examining the association of sociocultural factors with more behavioral data. Of course, this distinction between cultural norms and direct observation and experience is relative. First of all, both indices are based on self-reported data. Second, and most important, cultural socialization influences the perception of direct or observed experience. In this way, culture has an influence on what is an emotional situation, and on how to organize attention, recall and appraisals, among other factors (Moghaddam, 1998). However, reliability of observed behavior is usually higher than convergence on meaning. Anthropologists such as Mead and Freeman disagree on the sexual permissiveness of the Samoan society, but converge in sexual behavioral observation (Kottak, 1994). Moreover, some studies show that similar structural situations provoke the expected effect even in cultures that emphasize values contradictory to that situation. High work socio-structural position was associated with self-direction in Japan, as in the USA, including the case of Japanese women, even in a culture in which self-directedness for women has low value (Schooler, 1996).
f) Another criticism is the difficulty in establishing scale or metric equivalence. Cultural differences can be explained by differences in response style (acquiescence, moderation, social desirability). Studies usually correct set responses by using the individual general mean on the dependent variables as a covariable, or by using another form of intrasubject standardization. Thus, Schwartz (1992) recommended using the mean importance rating on values as a covariate in comparing multiple groups. On the other hand, Matsumoto et al. (1997) tested cultural differences using data standardized across individuals within each country (the findings were different than for the raw score analyses, and highly consistent across both values and behaviors). Results using intrasubject standardization or dichotomization of responses usually produce patterns of associations similar to the original scores. In other cases, scholars use relative scores (positive affect minus negative affect) that are not affected by different response styles. This criticism is, however, easy to check empirically (if it is true that Asian persons use more moderate step scales, they should show moderate means in all variables). This is not the general pattern. Thus, Chinese samples score lower in emotional intensity but relatively higher in the social desirability of negative emotions (see Basabe et al.), while Latin-Americans score lower in the expression of negative emotions but higher in the expression of positive emotions. Mayan subjects score lower on anger-related emotional reactions, but higher on fear/sadness-related emotional reactions (Martín et al.).
To conclude, the articles presented in this monograph help to advance knowledge of cross-cultural differences, whilst challenging some assumptions and showing, at least partly, how to overcome methodological criticisms and limitations.
We would like to thank the Projects (109.231 - HA 118/96; 109.231 - HA 208/97 and the University of the Basque Country group UPV 109.231 - G 56/98) and the 1998 Summer School organized by the University of the Basque Country, which helped us to contact collaborators and obtain valuable data. We are particularly grateful to E. Diener, J. Pennebaker, H. Wallbott and K. Scherer, who allowed us to use the original matrix data. Also we want to thank And Xu Lin Bao from the Beijing University.
Basabe, N., Páez, D., Valencia, J., González, J.L., Rimé, B., Pennebaker, J. & Diener, E. (1999). El Anclaje Sociocultural de la Experiencia Emocional de las Naciones: un análisis colectivo. Boletín de Psicología 62, 7-42.
Diener, E., Diener, M., & Diener, C. (1995). Factors predicting the subjective well-being of nations. Journal of Personality and Social Psychology, 69, 851-864.
Fiske, A.P.; Markus, H.R.; Kitayama, S. & Nisbett, R.E. (1998) The Cultural Matrix of Social Psychology. In D. Gilber, S. Fiske & G. Lindzey (Eds.). Handbook of Social Psychology, 4th edition. Boston: McGraw Hill.
Hofstede, G. (1991). Cultures and Organizations. Software of the mind. London: McGraw-Hill.
Hofstede, G. (1998). Masculinity and Femininity. The taboo dimension of national cultures. Thousands Oaks: Sage.
Inglehart, R. (1991). Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press.
Kenny, D. & La Voie, L (1985). Separating individual and group effects. Journal of Personality and Social Psychology, 48, 339-348.
Kottak, C.P. (1994). Antropología. Madrid: McGraw Hill.
Matsumoto, D., Weissman, M.D., Preston, K., Brown, B.R. & Kupperbusch, C. (1997). Context-specific measurement of individualism-collectivism on the individual level. The Individualism-collectivism interpersonal assessment inventory. Journal of cross cultural psychology, 28, 743-767.
McCauley, C., Ottati,V. & Lee, Y.T.(1999). National differences in economical growth: the role of personality and culture. In Y-T. Ling, C.R. McCauley & J.G. Draguns (Eds.), Personality and Perception across cultures. Mahaw: Lawrence Erlbaum Associates, Publishers.
Moghaddam, F.M. (1998). Social Psychology: exploring universals across cultures. New York: W.H. Freeman.
Páez, D., Martínez-Sanchez, F., Velasco, C., Mayordomo, S., Fernández, I. & Blanco, M.A. (1999). Validez psicométrica de la escala de alexitimia de Toronto (TAS-20): un estudio transcultural. Boletin de Psicología, 63, 55-76 [Psychometric validity of the Toronto alexithymia scale (TAS-20): a cross cultural study. Psychological Bulletin (Spanish)]
Páez, D., Ruiz, J.I., Gailly,O., Kornblit, A.L., Wiesenfeld, E. & Vidal, M.C. (1997). Clima emocional: su concepto y medición mediante una investigación transcultural. Revista de Psicología Social, 12, 79-89.[Emotional climate: its concept and measurement in a crosscultural research. (Spanish) Journal of Social Psychology].
Pandey, R., Mandal, M. K., Taylor, G. J. & Parker, J. D. A. (1996). Cross-cultural alexithymia: Development and validation of a hindi translation of the 20-item Toronto Alexithymia Scale. Journal of Clinical Psychology, 2, 173-176.
Peabody, D. (1999). Nationality Characteristics: dimensions for comparison. In Y-T. Ling, C.R. McCauley & J.G. Draguns (Eds.). Personality and Perception across cultures. Mahaw,NJ: Lawrence Erlbaum Associates, Publishers.
Pennebaker, J.W., Páez, D. & Rimé, B. (1997). Collective Memory of Political Events. Mahwah,NJ: Lawrence Erlbaum.
Pennebaker, J.W., Rimé, B. & Blankenship, V.E. (1996). Stereotypes of Emotional Expressivennes of Northerners and Southerners: A Cross-Cultural Test of Montesquieu’s Hypotheses. Journal of Personality and Social Psychology, 70, 372-380.
Sangster, R.L. & Reynolds, R.W. (1996). A test of Inglehart’s Socialization Hypothesis for the Acquisition of Materialist/Postmaterialist Values: The Influence of Childhood Poverty on Adult Values. Political Psychology, 2, 253-269.
Schooler, C. (1996). Cultural and social structural explanations of cross-national psychological differences. Annual Review of Sociology, 22, 323-349.
Schwartz, S.H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In M. Zanna (Ed.). Advances in experimental social psychology (Vol. 25, pp. 1-65). Orlando: Academic Press.
Schwartz, S.H. (1994). Studying human values. In A. Bouvy, F.J.R. van de Vijver, P. Boski & P. Schmitz (Eds.). Journeys into cross-cultural psychology. Lisse: Swets & Zeitlinger.
Wallbott, H. G. & Scherer, K. (1995). Cultural determinants in experiencing shame and guilt. In J.P. Tangney & K.W. Fischer (Eds.). Self-conscious emotions: shame, guilt, embarrasment and pride (p.465-487). New York: Guilford Press.
Aceptado el 20 de diciembre de 1999