Volume 7, No. 1, Art. 5 – January 2006

Further Explorations of the "Western Mind." Euro-American and German Mothers' and Grandmothers' Ethnotheories¹⁾

Heidi Keller & Carolin Demuth

1. Introduction

Ethnotheories about parenting and child development are mediating links between cultural "metamodels" (HARKNESS, SUPER, & VAN TIJEN, 2000) and parenting behavior. Cultural metamodels represent "a cluster of ideas that characterize cultures at a broad level and that should logically have wide-ranging functions for the organization of human development and social relationships" (p.23). They can be regarded as constituting filters through which the broader sociocultural orientations are implemented into parenting contexts and practices (SUPER & HARKNESS, 1996). The conceptions of independence and interdependence have been identified as such metamodels or general sociocultural orientations (FISKE, KITAYAMA, MARKUS, & NISBETT, 1998) specifying conceptions of the self and conceptions of relationships within particular eco-cultural environments. Independence denotes the self as separate from the social environment; interdependence denotes the self as intrinsically interconnected with the social environment. Independence conceives of the self as an autonomous agent, interdependence conceives of the self as a co-agent (GEERTZ, 1974; KAĞITÇIBAŞI, 1996a; MARKUS & KITAYAMA, 1991; SHWEDER & BOURNE, 1982). Although there is a broad consensus that some cultural environments can be described as oriented more towards independence and others as oriented more towards interdependence (FISKE et al., 1998; GREENFIELD & COOKING, 1994; KELLER & GREENFIELD 2000; ROTHBAUM, POTT, AZUMA, MIYAKE, & WEISZ, 2000), there is an ongoing debate about the nature, the structure and the boundaries of these conceptions (HARKNESS et al., 2000; KELLER, 2003; KILLEN, MCGLOTHLIN, & LEE-KIM, 2002; MILLER, 1997). [1]

We conceive of the cultural models of independence and interdependence as two independent dimensions representing systems of priorities (GREENFIELD, KELLER, FULIGNI, & MAYNARD, 2003; KELLER, YOVSI, & VÖLKER, 2002) that are instantiated both culturally and individually, thus containing substantial intercultural as well as intracultural variation. [2]

In this paper, we would like to concentrate on the independent sociocultural orientation which has often been equated with the Western conception of the person (GEERTZ, 1974; MARKUS & KITAYAMA, 1991) or the "Western mind" (GEERTZ, 1974; HARKNESS et al., 2000). The person is regarded as bounded, autonomous, self-confident, self-reliant, unique, competitive and curious. Relationships are conceived of as voluntary, negotiable and finite. Parenting can be understood as communicational exchange between two (quasi) equal partners (KELLER, 2003). It can be assumed that parents' socialization goals reflect these orientations, stressing the individual as an agent with own needs, feelings, desires and preferences from birth on (KELLER, 2003; KELLER, HENTSCHEL, YOVSI, ABELS, LAMM, & HAAS, 2004; LEVINE, 1988; SUPER & HARKNESS, 1996; WHITING & EDWARDS, 1988). In this paper, we want to analyze parental ethnotheories of mothers and grandmothers from two sociocultural environments that can be regarded as representing the Western independent sociocultural model, Euro-American urban middle class families in Los Angeles, USA, and Berlin, Germany (HOFSTEDE, 2001). There are not many studies analyzing similarities and differences in parental ethnotheories between Euro-American and European samples. The existing studies, however, reveal an interesting and rather consistent pattern that challenges the coherence and consistence of Western individualism and independence. In these studies which vary substantially in scope and methodology, Euro-American participants' attitudes to parenting appear to pursue straightforward individualism and independence, whereas the European participants' attitudes pursue independence as well as interdependence. Sara HARKNESS and collaborators (2000) have interviewed Dutch and Euro-American parents of small children with respect to descriptions and interpretations of their children's personality. The Dutch and the Euro-American parents both emphasized individualism, yet with a significant difference in meaning using different descriptors for individualism. The Dutch parents valued an enterprising and strong willed personality whereas the Euro-American parents valued a smart, self-confident personality and traits like being a leader. With respect to dependence, there was no agreement between the Dutch and the Euro-American parents. The Dutch parents saw dependence as a normal behavior of small children that they assumed to be innate. For the Euro-American parents, dependence was undesirable and rather interpreted as clingy. AUKRUST and collaborators (2001) analyzed Euro-American and Norwegian (as well as Turkish and Korean) parents' ethnotheories. They also found that Norwegian parents emphasized the importance of (long-term) close relationships, whereas the Euro-American parents had a more academic than a relational focus. Similarly, Swedish parents believe that the newborn and young child must be made to feel safe and secure within the home which symbolizes the strong attachment Swedes have to the family. Breastfeeding is seen as taking care of the child's emotional needs. Nevertheless, their socialization goals are individualistic, although traditional commandments with respect to self-inhibition and conformity are also existing (WELLES-NYSTRÖM, 1996). [3]

Comparisons of maternal beliefs of mothers in Southern Europe with those of Euro-American mothers revealed a similar pattern. NEW and RICHMAN (1996), for example, compared Euro-American and Italian participants' long-term and short-term socialization goals. While the Euro-American mothers first and foremost valued economic and emotional independence, Italian mothers primarily focused on the child's social relations. Short-term goals mentioned by the Euro-American mothers included a concern for the child's cognitive and emotional development and highlighted the mother's role in providing an environment rich in language with opportunities for exploration and play. Short-term goals of the Italian mothers were related to the care of an infant in terms of protection and assurance of health. TAMIS–LEMONDA, WANG, KOUTSOUVANOU and ALBRIGHT (2002) analyzed US-American and Greek (as well as Taiwanese) mothers' childrearing values with respect to 3 to 4 year old children. The Greek mothers emphasized decency (honesty/sincerity, responsibility) as well as education and independence, whereas Euro-American mothers emphasized self-maximization (assertiveness, creativity, curiosity) but at the same time values such as sharing and respecting others. [4]

Similarly, the attitude to teach children to conform to social norms in order to be "bien élevé" has been reported from several studies with French (Parisian) parents (SUIZZO, 2002) who, nonetheless, have also been described as oriented towards independence (RABAIN-JAMIN & SABEAU-JOUANNET, 1997). Studies differ though with respect to the importance of body contact (for a summary see SUIZZO, 2002). Also, developing good manners and placing importance on rules and respect for authority have been less emphasized by French parents (BAER, CURTIS, GRABB, & JOHNSTON, 1996; BORNSTEIN et al., 1996). Similarly, the Dutch parents in the study of HARKNESS and collaborators (2000) reported only a low profile for obedience. Obedience as a component of interdependence may be more adaptive in traditional hierarchical communities and not so much part of the conception of interdependence in Western societies (HARKNESS et al., 2000). [5]

Although these studies cover a variety of individualistic Western societies ranging from Northern to Southern Europe, there seems to be unexpected commonalities in contrast to the Euro-American society. European parents obviously emphasize interrelatedness as well as autonomous agency for their children, whereas Euro-American parents seem to focus primarily on autonomous agency. HARKNESS et al. (2000) as well as SUIZZO (2002) interpret their findings as supporting the conception of emotional interdependence that KAĞITÇIBAŞI (1996a) has proposed as a new synthesis of the dimensions underlying independence and interdependence: agency and interpersonal distance. The Euro-American mothers embody the model of independence, combining autonomous agency with separateness whereas the European mothers embody the model of autonomous relatedness, combining an autonomous agency with relatedness. [6]

The present study wants to further contribute to the understanding of parental ethnotheories as related to the sociocultural orientation of independence. We analyze Euro-American participants from Los Angeles, USA, and compare their views with those of participants from Berlin, Germany. Previous studies of parent-child interactions have revealed that German parents as well as US-American parents display primarily distal interactional behaviors that can be assumed to pursue independent socialization goals. Face-to-face exchange allows the infant to experience contingency, i.e. prompt responsiveness to own facial cues which supports the perception of the self as an independent agent (KELLER et al., 1999; KELLER, VÖLKER, & YOVSI, 2005). Object stimulation supports separateness and the development of technological skills (KELLER & GREENFIELD, 2000). Spending time on their own is regarded as fostering self-reliance (GROSSMANN & GROSSMANN, 1981). The display of those independence promoting behaviors has become significantly more pronounced in German middle class families (KELLER & LAMM, 2004), as well as in Euro-American families (DENCIK, 1989; ELDER, MODELL, & PARKE, 1993; GIDDENS, 1991) over the last decades. Both, Euro-American and German middle class mothers stress socialization goals that rely on independent cultural models (KELLER et al., under review) and, moreover, embody an independent orientation in their narrating styles (KELLER et al., 2004; KÄRTNER et al., 2004). [7]

Since parental ethnotheories can be regarded as expressing cultural metamodels, it can be assumed that they change in line with sociocultural changes over time within societies (LEVINE, MILLER, RICHMAN, & LEVINE, 1996; PALACIOS & MORENO, 1996; SUPER & HARKNESS, 2003; YOUNISS, 1995). In the Western world, an increasing individualism is recognized with respect to the globalization of capital, information, ideas and lifestyles describing the post-industrial society. These changes can be assumed to also inform parental socialization goals and parenting ideas and practices towards an increase in orientation towards the model of independence (GIDDENS, 1991; KELLER et al., 2003; KELLER & LAMM, 2004). Mothers of small children and the grandmothers of these children thus can be assumed to have been raised in different sociocultural environments. In the present study, we therefore also analyze grandmothers' socialization goals and ethnotheories and compare them to those of mothers of three months old infants. [8]

The present paper follows two objectives: 1. Examining the contents of parental ethnotheories in Euro-American and German middle class families with the assumption of a more independent orientation of the Los Angeles participants and an independent as well as interdependent orientation of the Berlin participants; 2. Examining intergenerational changes in parental ethnotheories with the assumption of an increase of the independent orientation in the younger generation in the Euro-American as well as in the German samples. [9]

2. Method

2.1 Participants

Participants of this study were mothers of three months old infants and their mothers living in Los Angeles, CA, and Berlin, Germany. The families all belong to an urban educated middle class (18 of the German mothers and 5 of the German grandmothers had at least the entrance qualification for the university ("Abitur"); all the Los Angeles mothers and all the Los Angeles grandmothers except one hold at least a bachelor's degree). The samples reported here form sub-samples of a larger study of N = 25 mothers in the Los Angeles and N = 41 mothers in the Berlin sample. Criterion for inclusion in this study was that assessments with mothers and grandmothers were possible. The socio-demographic information of the participants is reported in Table 1.

Table 1: Sociodemographic information for the different samples [10]

The differences in sample size (21 mother-grandmother pairs in Berlin and 7 mother-grandmother pairs in Los Angeles) are mainly due to the local distance of their places of residence. The grandmothers of the Los Angeles sample were living all across USA and thus in part not accessible for the study. Only those mothers were included in the study where both, the mother and the grandmother were available. Considering the resultant small size of the grandmother sample, the present findings with respect to the grandmothers will be regarded as exploratory. However, the sociodemographic characteristics of the Los Angeles mothers included in this study are not different from the rest of the Los Angeles sample where the grandmothers were not available. Also, further analysis of the ethnotheories of these two sub samples (mothers and grandmothers) did not reveal differences in their ethnotheories (NETTE, 2003), so that they can be regarded as representative of Euro-American middle class families. This approach capitalizes on prototypical cultural samples that are defined by sociodemographic characteristics. The age of the mothers and grandmothers as well as of their partners and the birth rank of their children are comparable across the samples. Differences concern the gender composition of the samples, the marital status of the mothers, and the participation in the labor force of mothers and grandmothers. [11]

2.2 Procedure²⁾

2.2.1 Sampling

The Los Angeles sample was recruited between mid July and mid October 2002. Participants were contacted through the distribution of information material in baby classes, clinics, as well as through personal recruitment and snowball technique. The appointment for the assessment was set by telephone. US-American and (bilingual) German research students made home visits when the baby was between 11 and 14 weeks of age in order to collect the data. After completing the data collection, mothers were asked whether their mothers lived in Los Angeles or nearby and whether they would be willing to participate. An extra appointment was then set with the grandmother who was also interviewed in her home. Participation was completely voluntary on the basis of informed consent³⁾. [12]

The German sample was recruited between September 2003 and January 2003. Participants were recruited through a community hospital. Pregnant women and women with very small babies were informed about the study through distribution of information material. Interested mothers could sign up and were then contacted by research assistants. Families were also recruited through snowball technique. The procedure of contacting families was equivalent to the Los Angeles procedure⁴⁾. All participants agreed to take part in the study on a voluntary and informed consent basis. No reimbursement was given to the participants. [13]

2.2.2 Data collection

Although qualitative methodology is meanwhile becoming established in the field of psychology, qualitative methods are still rarely used in empirical studies within the realm of developmental psychology, and a stronger and more equal use of a qualitative methods in this field has been called for (MEY, 2000, 2003). In our present study, we therefore decided for a twofold methodological approach and used two types of measurement: a questionnaire specifying socialization goals for the first year of life (see also Table 2) and a qualitative semi-structured picture card interview based on our own previous work on parental ethnotheories (BORKE, 2002; EICKHORST, 2002; KELLER, YOVSI, & VÖLKER, 2002; KELLER, VOELKER & YOVSI, 2005). In addition, sociodemographic data was collected. This method triangulation (e.g., FLICK, 1995) was used to systematically enhance and complement the findings. [14]

The home visits were structured as follows: One research assistant visited the family at a time that was indicated as convenient for the mother. After a warming-up phase, socio-demographic information was assessed and the participant was asked to answer the questionnaire. Then, the picture card interview was conducted. [15]

Socialization goals questionnaire. Participants answered 18 statements, embodying socialization goals for the first three years of children's lives with a 1 – 6 point rating scale (1 = do not agree at all; 6 = agree completely). The items were compiled on the basis of literature reviews on the definition of independence and interdependence (e.g., KAĞITÇIBAŞI, 1996a). Nine statements represent independent socialization goals: (1) develop joy in life, (2) develop independence, (3) develop self-confidence, (4) become assertive, (5) develop the ability of enforcement, (6) develop a sense of self-esteem, (7) develop creativity, (8) develop competitiveness, (9) develop a sense of self; 9 statements represent interdependent socialization goals: (1) learn to control emotions, (2) develop close personal relationships, (3) learn to understand the emotions of others, (4) learn to obey the parents, (5) learn to obey elderly persons, (6) learn to care for the well-being of others, (7) learn to help others (mother, siblings), (8) learn to cheer up others, (9) develop attachment to the family. [16]

Picture card interview. At the beginning of the interview, it was stressed that we consider women as experts of child development and that we are interested in women's opinions about baby care in different cultures. It was emphasized that there are no right or wrong answers and that we are interested in the personal opinion of the particular woman. The mother was presented 5 picture cards (21cm x 15cm) demonstrating five parenting systems as suggested by the Component Model of Parenting (KELLER, 2000, 2002) covering the following infants caretaking experiences: (1) breastfeeding/nursing, (2) body contact, (3) body stimulation, (4) object stimulation, and (5) face-to-face contact. The pictures were taken from mother-infant interactions from the same sociocultural group the mother belonged to (Berlin middle class and Los Angeles middle class mothers). The mother was asked to pick the card which represented best maternal care for a three months old baby from her point of view. The other cards were removed and the mother was asked why she considered the selected card as best care and what was important about it in her opinion. She was probed until she did not add any new information. The same procedure was repeated until only one card was left. With the last card, the mother was asked why she picked this one last. The procedure was equivalent for the grandmothers. The picture cards served as narration inciting stimuli, therefore the ranking of the picture cards itself was not considered in the later analysis. All interviews were tape recorded and fully transcribed afterwards. For the sake of better legibility, the interview excerpts quoted in the following do not include any pause symbols (*) although they were used in the original transcription. The interview procedure (questionnaires not included) lasted about 20 to 30 minutes. [17]

2.2.3 Data analysis

Socialization goals questionnaire. The mean scores for the independent and the interdependent scale were computed (see Table 2). [18]

Picture Card interview. Data analysis was accomplished by inductively identifying topics that the participants addressed in regard to parental ethnotheories (cf. HARKNESS et al., 2000; MILES & HUBERMAN, 1994, for similar procedures). Following the principles of qualitative text analysis, the aim was to develop categories as close as possible to the verbal material. We used a technique that borrows in large from qualitative content analysis (e.g., MAYRING, 2000a, 2000b). Qualitative content analysis aims to preserve the advantages of quantitative content analysis for a more qualitative text interpretation, e.g. by observing rules of analysis (the material is to be analyzed step by step, following systematic rules of procedure) and by combining inductive category development with deductive category application. Categories were carefully derived from the text material and checked against the data within the process of analysis. In the first step of the procedure, criteria of definition derived from the theoretical background of the Component Model of Parenting, as well as the present research question were formulated in order to determine which aspects of the textual material should be taken into account. Following these criteria, the material was worked through and categories were step by step inductively derived. Examples from the text for each category were discussed with a group of three researchers trained in checking for reliability. 90% of agreement were reached. In the next step, the derived categories were applied to the text material. [19]

In this way, main topics in regard to parental ethnotheories were identified in the verbal material. Topics were identified independent of the pictures, i.e. if a participant talked about the same topic with respect to different picture cards; it became part of the same narration unit. The position of a topic is equivalent with the first mentioning. In a second step, the interview data was systematically analyzed with respect to each of the identified topics for each participant. The relevant text passages were analyzed in regard to the way these topics are described by the individuals. Descriptors for each topic were identified and listed as codes under the relevant category (topic) (e.g. holding close, body contact, cuddling etc. were listed under the category "body contact"). Developmental consequences were coded separately. In a final step, descriptors belonging to the same functional category were compiled, e.g. "food, "breast milk," "nutrition," "health," "immune system" were coded as "health." Furthermore, quantitative aspects were included in the analysis by taking into account the frequencies of coded descriptors. The lists of descriptors per topic are the basis of the presented analysis. Moreover, the frequency of descriptors per topic was counted and is indicated in brackets in the result section. Frequency of mentioning a topic was interpreted as indicator for the subjective importance of the topic to the mother. Because of the compilation, the frequencies of descriptors may exceed the number of participants. [20]

3. Results

3.1 Socialization goals questionnaire

Table 2 shows the mean and standard deviations of the independent and interdependent scales according to generation and region.

Table 2: Socialization goals for the first three years of life (1 = do not agree at all, 6 = agree completely) [21]

Other than expected, there are no significant differences between the samples and between the generations. Independent and interdependent socialization goals are selected to similar degrees by all participants. Yet there are slight tendencies for higher independence scores as compared to the interdependence scores across all samples. There are also tendencies for higher interdependence scores for grandmothers in both samples, which supports our second assumption of an increase of independent orientation in the younger generation. Given the relative small sample size, complementary results from the analysis of the picture card interviews will be taken into account in order to further substantiate our findings. [22]

3.2 Picture card interviews

The numbers of topics mentioned in the four samples are listed in Table 3.

Table 3: Number of topics mentioned by the participants [23]

The LA mothers and grandmothers address fewer topics than the Berlin mothers and grandmothers. Mothers and grandmothers within the same culture, however, do not differ from each other. The following topics were identified for all samples: (1) breastfeeding, (2) body contact, (3) playing/Beschäftigung, and (4) avoiding overstimulation. Only the Berlin grandmothers addressed one more topic: importance of the father. This topic is, however, not included in the further analysis. [24]

a) Preference of topics: the participants from the different samples address similar topics with similar priorities as is indicated in Table 4. Breastfeeding and nursing has the highest priority for all participants, followed by body contact and close (bodily) proximity. Play and stimulation are the next frequent mentioned category. The German Beschäftigung is even almost equally frequent as close proximity. The next frequent aspect concerns the avoidance of mainly overstimulation, especially with respect to motor development (cf. Table 4).

Table 4: Preference of topics (mean rank) [25]

b) Descriptions of the four common topics: (The numbers in parentheses indicate the frequency of descriptors mentioned per topic). [26]

3.2.1 Breastfeeding/nursing

The Berlin mothers. The main importance of breastfeeding for the Berlin mothers lies in the intimate and intensive mother-child situation, characterized by close body contact (11). The possibility for eye-contact is appreciated as well (2). The second important aspect is seen in the healthy nutrition (7). Breastfeeding, on the one hand, is simply seen as a necessity to survive and as a very practical solution to feed the child; on the other hand, it is considered to provide the infant with love, security and confidence which might not be possible to the same extent when feeding the child with a bottle. These are the important cornerstones for a good mother-child relationship that have a positive impact on development and well-being of the child. This bodily and emotional closeness is also believed to contribute to the development of self-confidence in the child.

Example:

"Weil Stillen und die körperliche Nähe für das Kind ganz wichtig sind […] Also die körperliche Nähe für die geistige und körperliche Entwicklung, um Körpererfahrung zu kriegen und Liebe zu erfahren. Und das Stillen, weil Muttermilch nun mal für Babys sinnvoll ist."

"Because breastfeeding and bodily proximity are really important for the child. […] I mean for the mental and physical development to experience body sensations and love. And breastfeeding [is important] because breast milk is simply good for babies." (Berlin mother 08, translated by the authors) [27]

The Berlin grandmothers. The Berlin grandmothers likewise value the intimate situation, the heartfelt relationship and the body contact between mother and baby as the most important benefits of breastfeeding (11). They too appreciate the nutritional value as a basis for healthy development (4). For them, breastfeeding lays the foundation for love, security and con