Commentary for “Parenting Culture(s): Ideal Parent Beliefs Across 37 Countries”

Culture Areas

The idea of culture areas is inherently appealing as an aid to building expectations of the thinking and behavior of people from different parts of the world. As researchers, designers of social or health interventions, or even teachers, familiarity with one part of a culture area can often help us make more educated guesses about what we will learn when working with people from another part of that area. This apparently simple approach is full of pitfalls, however.

For starters, designating who or what places belong to which culture areas has been a continuing challenge to anthropologists since the earliest days of the field. George Peter Murdock, founder of the Human Relations Area Files (HRAF) at Yale University, produced a map of 11 major clusters and 48 sub-clusters in Africa, based on published ethnographic information including history, geography, language, economy, and a variety of customs (Murdock, 1959). Other efforts at creating culture areas were made for native American groups in North America and Oceania (Burton & White, 1991). The idea of culture areas was also of concern to anthropologists in pursuit of explanations about why certain customs or beliefs co-occur; in this case, though, anthropologists worried about “Galton’s problem,” namely that the co-occurrence of particular practices might be due to their membership in the same culture area rather than causally related through psychological or social mechanisms. Summarizing the history of anthropological efforts to identify culture areas, Carol Ember, the current director of the HRAF, comments, “The long and short of it is that I wouldn’t recommend emphasizing culture areas at all” (personal communication).

This conclusion is logical, yet frustrating. How can we understand different cultures—much less work productively with people from other cultural places—if we are not able to create a sense of order among them? One currently popular solution to this dilemma in cross-cultural psychology is the “individualism/collectivism” dichotomy defined originally by Hofstede and elaborated by Hazel Markus and Shinobu Kitayama (1991), building from contrasts they perceived between their own respective cultural places, namely middle America and Japan. Although the I/C dichotomy was not directly proposed as a way to designate global culture areas (two very large ones!), it has often been translated to that effect. There are problems inherent in this application, however: in particular, there are tremendous differences among societies within each of those categories which the I/C dichotomy cannot capture, as the authors of this paper note.

The concept of culture areas in both anthropology and psychology relies on several kinds of information, in contrast to the approach taken in this paper. How well does this very restricted focus work? The answer is that creating culture areas based on just one tiny piece of the culture puzzle works surprisingly well—but it depends on just how narrowly culture areas are defined. Leaving aside their analyses of educational differences within culture areas, the authors propose five parenting culture zones and their country composition. At first glance, some of the areas seem reasonable: for example, the Asian culture area does include three east or south Asian countries (Thailand, China, and Japan) in addition to Iran and Turkey, the African area includes all the African samples (Egypt, Burundi, Algeria, Togo, and Cameroun), and the Hispanic-Italian area includes all the Hispanic countries, plus Italy at the very edge of the area. The two Western culture areas are more problematic, as they group together eastern European countries (Poland, Romania, and Serbia), Russia, north-western European countries (Germany, Austria, The Netherlands, and Finland), and the British diaspora (Australia, the United States, and the United Kingdom)—a broad category that includes a great deal of internal variability. Even more perplexing is the “Western II” culture area, consisting of Canada, Belgium, Switzerland, France, and Lebanon.

The explanation for this last category is probably linguistic: all the samples in the Western II group (except Lebanon) were drawn from French-speaking areas—including Quebec in Canada, and the French-speaking section of Switzerland. This should serve as a reminder that cross-cultural analyses should not be confused with cross-national ones. The question still remains, however, of why the Western II sites were differentiated from the other Euro-American and British diaspora sites—or why the Western I culture area includes such a diverse set of countries. The answer to these questions seems to be due at least in part to the method used to create the culture areas. Thanks to the first author, who has shared with me the results of a second method (document clustering), we can see associations among the different samples at various levels, from pairs or triplets of countries most similar to each other, up through consecutive groupings of larger clusters. The lowest level groupings in this solution generally (although not always) correspond more closely to groupings that one might expect based on cultural heritage and language. For example, the British diaspora samples are associated with each other at the lowest level, as are the north-western European samples, while Russia is most closely associated with Poland. This method also differentiates an Islamic cluster including Lebanon, Algeria, Iran, Egypt, and Turkey, plus Serbia (which includes some Islamic heritage). The document clustering method is less successful, however, in keeping together the Hispanic samples, which are spread among various mid-level clusters. In summary, it appears that language may be the strongest determinant of culture area assignments—perhaps relevant to the fact that the data used for classification is linguistic. To consider how useful these culture areas are, we turn to the commonalities and differences in the meanings associated with them.

Cultural Themes

A central goal of this paper is to encourage researchers in the behavioral sciences to take a “bottom-up” approach to discover patterns of cross-cultural variability, rather than trying to fit individual cultural groups into pre-existing categories. The method they used to analyze individual one-word descriptors of the “ideal parent” in their multi-country study is, logically, a linguistic program, Semantic Network Analysis (SNA). As the authors succinctly describe this approach, “. . .the SNA proceeds according to a principle of increasing abstraction, in which the responses provided by the respondents . . . are organized into concepts that in turn are organized into themes (i.e., clusters of concepts)” (REF page). I found some aspects of this approach confusing—for example, the fact that some characteristics of the ideal parent end up being coded as nouns rather than adjectives, and the appearance of the same words in both the themes and their component concepts, as listed in Table 2 in article Parenting Culture(s): Ideal-Parent Beliefs Across 37 Countries (URL of the article will be included here).

In general, however, the results are quite clear: as the authors state, the theme of being “loving” was found to be “the core of ideal-parent beliefs” in the three Western culture zones (Western I and II, and Hispanic-Italian). Both the Asian and African culture zones have their own distinctive central themes, with “responsibility” and “respect” characterizing the African zone while “family” plus “responsibility” are central for the Asian culture zone. These last two profiles are consistent with research literature in both places. For example, much research has documented the emphasis on training children to be responsible and cooperative contributors to the family unit in sub-Sarahan communities (Harkness et al., 2015; Serpell, 2011; Super & Harkness, 1986; Tsamaase et al., 2020). Likewise, research on Asian families consistently portrays a primary focus on inter-generational commitment via the Confucian principle of “filial piety” (Ho, 1996) and more generally in the context of “collectivism.”

Does this mean that African and Asian parents don’t love their children? Obviously it does not. Rather, in both kinds of families, love is implicit, expressed even in small gestures such as, in Asian families, cooking a child’s favorite food or staying up late to supervise homework. Among the Yoruba of southern Africa, Zeitlin (1996) describes fathers voluntarily sharing their own allocation of food with their children. Even physical punishment, in this cultural context, is thought of as an act of kindness (Zeitlin, 1996).

In contrast to the interpretable results of the African and Asian culture areas, the shared theme of love in the three Western cultures seems less satisfactory in that it fails to capture real differences in the ways that parents express love, or perhaps even experience it. For example, the concept of “familismo”—dedication to the family’s cohesion and well-being—does not emerge in the Hispanic-Italian cultural group. Differences in emotional expressiveness that are often noted in northern versus southern European societies are not evident here; and the contrast between “empathic” (Western I) and “understanding” (Western II) seems more like a linguistic than truly psychological distinction.

Concluding Thoughts

This paper reports on an impressively large-scale research project. The researchers deserve congratulations for carrying out what must have been a major effort, requiring close collaboration across many countries. The focus in this paper on a “bottom-up” approach is also welcome as a counterpoint to more typical approaches in cross-cultural psychology that employ pre-existing categories to explain cultural differences—usually with limited success. The fact that one tiny piece of the cultural puzzle from each research site can produce any meaningful results is also impressive. At the same time, this study demonstrates the limitations of such an approach. If we really want to understand the whole puzzle, as well as create locally meaningful interventions to support families, more different kinds of data are essential. Culture is indeed a puzzle in both senses discussed here, and a complex one at that, but the effort reported in this paper is a good start.