Early Learning Environments for the Development of Attention: Maternal Narratives in the United States and Japan

Abstract

A large body of research has demonstrated cross-cultural differences in visual attention, especially between members of North American societies (e.g., Canada, United States) and East Asian societies (e.g., China, Japan, Korea). Despite an increasing number of studies suggesting an emergence of cross-cultural differences in early childhood, relatively little is known about how these culturally divergent patterns of attention are acquired and maintained. It has been largely assumed that socialization practices, especially parent–child interactions, contribute to the acquisition of cross-cultural differences in attention. By focusing on maternal narratives during the shared reading activity, this study examined the socialization contexts in which mothers direct their infants’ attention in the United States (n = 50 dyads) and Japan (n = 53 dyads). Mothers in the United States and Japan read a picture book to their 6- to 18-month-old infants in the lab, and maternal narratives were coded to identify attention to focal objects and social interactions. Infants’ sustained attention was also measured during shared reading. The findings demonstrated that during the shared reading activity, U.S. mothers were relatively more likely to focus on the focal objects than the background, whereas Japanese mothers were more likely to refer to the social interactions between focal objects and the background. Infants’ age and gender were not related to maternal narratives, and infants’ sustained attention was similar across cultures. Findings suggest significant cross-cultural differences in mother–infant interactions, which may act as scaffolds for infants to internalize their parents’ cognitive styles.

We are surrounded by endless amounts of visual information. While we are awake, our mind is continuously receiving information. To efficiently process and perceive visual information, it is necessary to attend to specific areas of interest. A wealth of empirical evidence has documented systematic and significant cross-cultural differences in attention, particularly comparing those from North American and East Asian cultures (for a review, see Masuda, 2017). It has been largely assumed that socialization practices, especially caregiver–child interactions, contribute to the acquisition and maintenance of cross-cultural differences in attention; however, little is known about the actual process of this cultural transmission. By using an ecologically sensitive method to investigate mother–infant interactions in the United States and Japan, the current research sought to examine early socialization practices related to the development of visual attention.

Visual Attention in the United States and Japan

According to the theoretical model of analytic versus holistic cognition, the ways people allocate their attention are different across cultures (Nisbett, 2003; Nisbett & Masuda, 2003; Nisbett & Miyamoto, 2005; Nisbett et al., 2001). Cultural differences in patterns of attention reflect philosophical, ecological, agricultural, and other civilization practices uniquely different across cultural groups (Nisbett, 2003; Uskul et al., 2008; Varnum et al., 2010). Members of North American and Western cultures (e.g., Canada, England, the United States) tend to exhibit an analytic pattern of attention (e.g., Nisbett et al., 2001; Nisbett & Masuda, 2003; Nisbett & Miyamoto, 2005). The analytic pattern of attention is characterized as focused attention on the focal object(s) in isolation from the background. Nisbett and colleagues suggest that this pattern of attention reflects a worldview that echoes Ancient Greek traditions, which tend to identify individual objects as separate entities with unique properties that are detached from the contextual information. In contrast, attentional styles commonly shared by members of East Asian cultures (e.g., China, Japan, Korea, and Taiwan) are more holistic. The holistic pattern of attention is characterized as attention to both focal objects and contextual information, as well as to relationships among objects. This pattern of attention is linked to cognitive holism, which refers to the idea that the part cannot be understood without understanding its relation to the whole (Choi & Choi, 2002; Ji et al., 2001; Peng & Nisbett, 1999; Spencer-Rodgers et al., 2009). This thought system is rooted in the traditional religions and beliefs shared in East Asian regions, such as Taoism, Confucianism, and Buddhism.

A number of empirical investigations have provided supportive evidence for the theoretical model of analytic versus holistic cognition. For example, Masuda and Nisbett (2001) developed the Michigan Fish task and asked college students from the United States and Japan to describe short vignettes of underwater scenes. U.S. students were more likely than Japanese students to focus on the main objects in their verbal description. Japanese participants were more likely than U.S. participants to describe the contextual information, as well as relationships among objects. Similarly, in a study that examined the performance of the Framed Line Test (FLT; Kitayama et al., 2003), the U.S. participants were better than the Japanese at the absolute task, which required selective attention focusing on the target line and ignoring the contextual information. Conversely, Japanese participants were better at the relative task, which required participants to incorporate contextual information.

Recently, some researchers have suggested that cultural differences in social orientation (independence vs. interdependence) and relational mobility (high vs. low) may facilitate culturally divergent patterns of attention (e.g., Markus & Kitayama, 1991; San Martin et al., 2019; Varnum et al., 2010). Indeed, previous studies have revealed that for both adults and children, the social nature of the stimuli particularly enhances cross-cultural differences in attention (Masuda et al., 2008; Senzaki et al., 2018). The U.S. culture tends to endorse and afford independence, autonomy, and self-expression, and selective attention may be shaped by these cultural values. Given these traditions and cultural practices prevalent in the United States, it may be important for parents to instill their children with the analytic pattern of attention. On the other hand, people in Japan tend to experience social contexts in which members of a given group are relatively fixed and opportunities to develop new relationships are limited (San Martin et al., 2019). In such environments, it is important to preserve social harmony with their in-group members; thus, it may be more important for people in Japan to broaden their attention to details in their social behaviors and the surrounding context.

There is a growing interest in investigating cross-cultural differences in attention among children. Several studies have suggested that cross-cultural differences in children’s attention emerge at around school age. For example, 6- to 13-year-old Japanese children exhibited a more holistic pattern of attention when they completed the FLT than the U.S. children (Duffy et al., 2009). Similarly, Senzaki et al. (2014b) used a drawing task and found that the effect of cultural differences in attention increased from the first to the sixth grades. Other studies indicate that cultural differences in attention increase gradually during childhood. For example, Imada and colleagues (Imada et al., 2013) demonstrated a gradual increase of cross-cultural differences in attention from ages 4 to 9. In this study, children were asked to complete a scene description task (similar to the Michigan Fish task) and the Ebbinghaus illusion task, which uses an optical illusion that occurs when objects (e.g., two circles of identical size) are presented with surrounding circles of smaller or larger sizes. Compared with the U.S. children, Japanese children were significantly more likely to mention the background information and discuss social interaction. Japanese children were also more susceptible to the illusion, and cross-cultural differences in attention increased with children’s age.

More recent studies have reported significant cross-cultural differences even in early childhood. Kuwabara and colleagues found that U.S. children outperformed Japanese children as young as at 3 years of age on a task that required object-oriented selective attention, whereas Japanese children outperformed U.S. children on a task that required relational understanding among objects (Kuwabara & Smith, 2012, 2016). In their study, children were shown some representative features of an object (e.g., ears, tail, and legs of a dog), and children were asked to identify the object based on these features. The U.S. children were better able to identify the object, indicating that they tend to use object-oriented selective attention. These studies suggest that early socialization practices, such as parent–infant interactions, may play an important role in developing culturally shaped patterns of attention.

Although cultural differences in attention are attributed to different early socialization practices (Keller, 2003; Markus & Kitayama, 1991; Miyamoto & Wilken, 2013), little empirical work has examined early socialization practices as a foundation for cross-cultural differences in attention. Developmental theories suggest that cognitive development is embedded in the social world. Early social interactions, especially caregiver–child interactions starting in infancy, may play a key role in shaping children’s development of attention (e.g., Gauvain, 2002; Keller, 2017; Rogoff et al., 1993; Vygotsky, 1930/1978; Yu & Smith, 2016). Through social interaction, children and their caregivers “extract meanings, assign interpretations, and infer intentions” (Bruner, 1983, p. 29).

Socialization as Vehicle for Cultural Transmission

Although socialization practices can take a variety of forms, narrative serves as an important cultural tool in organizing information in a meaningful way. Narrative is the process of constructing one’s thoughts in the form of storytelling, and a previous study has shown that narrative plays an important role in adults’ attention allocation across cultures (Senzaki et al., 2014a). Narrative construction is the active practice of meaning-making that human children are exposed to from birth, and it helps them understand their physical and social surroundings through cultural meaning systems (Bruner, 1990; Fivush, 2009; Nelson, 2009); therefore, we consider narrative construction as the vehicle of cultural transmission that shapes children’s development of attention.

There are some investigations which focused on cross-cultural differences in narrative construction during parent–child interaction (e.g., Wang, 2006, 2013; Wang & Brockmeier, 2002; Wang et al., 2017; Wang & Song, 2018). For instance, Doan and Wang (2010) found that European American and Chinese immigrant mothers used different narratives when they created a story while reading a picture book to their children. Moreover, a study conducted by Senzaki and colleagues (2016) used the joint Michigan Fish task, and the results showed that Canadian parents were significantly more likely to discuss the main objects with their 4- to 9-year-old children than Japanese parents, while Japanese parents were more likely to discuss the contextual information with their children than Canadian parents.

Even before children start to develop their language skills, their caregivers have opportunities to begin directing infants’ attention naturally and spontaneously through daily activities. Indeed, cross-cultural differences in caregiver–child interactions have been documented in infancy (for a review, see Rothbaum et al., 2003). One of the particularly important studies was conducted by Fernald and Morikawa (1993), who tested how U.S. and Japanese mothers talked to their 6- to 19-month-old infants. In their study, mothers and infants were observed at their home while they played with two sets of target toys (dog and pig stuffed animals; car and truck wooden vehicles). Their observations showed that U.S. mothers were more likely to label the target objects (e.g., “This is a car”), whereas Japanese mothers were more likely to engage their children in social interaction play (e.g., “Mr. Doggy said hello”). This study provides partial evidence for object-oriented attention in the U.S. mothers; however, implications for the development of attention are limited as the study was originally designed to examine different linguistic inputs, especially focusing on mothers’ noun labeling.

Taken together, there are theoretical and empirical foundations from the developmental perspective that suggest mother–infant interactions play a critical role in understanding cross-cultural differences in attention. Despite the convincing evidence for cross-cultural differences in visual attention between the U.S. and Japanese adults and children, studies examining the developmental source of cultural differences are limited. Because cross-cultural differences in attention have been found in early childhood, it is important to examine early socialization environments in different cultural contexts. By examining caregiver–infant interactions in the United States and Japan, we sought to shed light on the process of cultural transmission whereby infants acquire patterns of attention shared by their parents’ cultural groups.

Current Study

Building upon previous research from both cultural and developmental psychology, we hypothesized that culturally unique socialization practices start early in infancy. To capture the socialization of analytic–holistic attention in the United States and Japan, the present study tested mother–infant interactions while mothers read a picture book to their 6- to 18-month-old infants. This activity was selected because it involves narrative construction (i.e., storytelling), and it is part of typical childrearing practice for families in both the United States and Japan (Ishizaki, 1996; Murase, 2014; Toyama, 1989). Previous research has used a similar task to investigate parent–child socialization practices across cultures (Doan & Wang, 2010; Senzaki et al., 2016). We focused on 6- to 18-month-old infants for two reasons. First, this is the period infants begin tuning into their caregivers’ attention, and parent–infant interactions become the foundation to establish further social and cultural learning (Bruner, 1995; Butterworth, 2004). Second, the beginnings of cross-cultural differences in attention have been observed as early as 24 months (Waxman et al., 2016); therefore, this study aimed at examining early socialization practices preceding cultural differences in children’s attention.

We predicted that narratives created during the shared reading task would reflect the process by which mothers encourage their infants to develop specific patterns of attention that fit with their cultural emphases. Based on past research (Masuda & Nisbett, 2001; Senzaki et al., 2014a, 2016), we coded utterances created by mothers during shared reading as a reflection of analytic–holistic attention. An utterance was defined as the smallest meaningful unit. We identified the focal objects and the background information in the book and assessed how often mothers referred to the focal objects and the background, as well as social interactions among the objects. We hypothesized that the U.S. mothers would be more likely to direct their infants’ attention to the focal objects, whereas Japanese mothers would be more likely to direct their infants’ attention to social interactions between focal objects and the background. In addition to culture, we included infants’ gender as a factor because previous research has indicated that infants’ gender may play a role in the ways mothers talk to them (e.g., Fivush, 1989; Fivush & Wang, 2005).

Although we predicted that maternal socialization practices would differ across cultures, we expected that 6- to 18-month-old infants’ development of sustained attention would be similar in the United States and Japan. Previous research suggests that at 24 months of age, infants in the United States and China show overall similar patterns of attention (Waxman et al., 2016), and even Japanese infants as young as 6 to 8 months of age have demonstrated object-based attention (Tsurumi et al., 2018). We thus predicted that infants’ sustained attention would be similar between U.S. and Japanese 6- to 8-month-old infants.

Method

Participants

Participants included 103 mother–infant dyads (50 U.S. and 53 Japanese). The average age of the U.S. infants (31 boys and 19 girls) was 10.44 months (SD = 3.60, range = 6–18 months), and the average age of Japanese infants (34 boys and 19 girls) was 9.98 months (SD = 3.95, range = 6–18 months). Participating infants’ age and the gender distribution were not different across cultures, t(101) = 0.62, p = .54, and χ²(1) = 0.05, p = .82, respectively. As previous research has demonstrated that maternal education level can influence socialization practices (e.g., Chavajay & Rogoff, 2002), we recruited U.S. and Japanese mothers with similar educational backgrounds. Maternal education level was similar across cultures, χ²(5) = 7.43, p = .19, as 78% of the U.S. mothers and 83% of Japanese mothers had attained at least a college degree. Families in the United States were recruited in Green Bay, Wisconsin, and Japanese families were recruited in Saitama, Japan. Infants and mothers came to the laboratory as part of an ongoing investigation focused on early social and cognitive development, which was reported elsewhere (Shimizu et al., 2018). All the U.S. infants and mothers were born in the United States and spoke English as their first language. Based on maternal report of the U.S. children, 43 children (86%) were identified as White or European American and seven children were identified as mixed race/ethnicity. All Japanese infants and mothers were born in Japan and spoke Japanese as their first language.

Procedure

Signed consent forms from the mothers were submitted upon their arrival to the research laboratories. Testing was completed at the child development laboratories at local universities in the United States and Japan. The institutional review board at local universities in the United States and Japan approved study procedures. The Japanese instructions were translated and back-translated from English to Japanese. Families in the United States received a small toy or book as a gift for the infant, and Japanese families received a gift card to cover the transportation cost.

Mother–infant reading task.

Ten Little Ladybugs by Melanie Gerth and Huliska-Beith (2001) was selected as the stimulus because it included a clear description of focal objects (ladybugs) and the background information. Two children’s literacy experts (one from the United States and one from Japan) confirmed this as an age-appropriate book for the current sample of infants in both cultural groups. To encourage mothers to spontaneously talk about the picture, the storylines were masked with colored-tape so that only pictures were available to talk about. The book starts with 10 tactile ladybugs, and one ladybug disappears on each subsequent page. On each page, there are ladybugs, a critter (e.g., a bird, bees, butterflies, etc.), and other background information, as well as a numeric sign showing the number of ladybugs. The book has a total of 11 pages. Mothers first flipped through all pages by themselves to think about a story, and then they were instructed to read the book to their infants. They were told “Please read this book to your child in the way you would at home. You can make up a story in any way you would like.” As a manipulation check, mothers were asked what they thought the main characters were after reading the book, and all mothers in both cultures indicated that they thought the ladybugs were the main characters. The mother–infant interaction was videotaped and later transcribed verbatim to code maternal narratives.

Sustained attention.

Infants’ sustained attention was measured during shared reading. Based on previous research, longer looking time was used as an indicator of sustained attention due to the complexity of the stimuli and the age of infants (Graziano et al., 2011). Two coders who were blind to the hypothesis coded the overall duration of time infants looked at the book (in seconds), and the intercoder reliability for 25% of the data was κ = .92. An open-source program, CowLog (Matti, 2016) was used for encoding the data.

Coding of Maternal Narratives

Coding was performed in the original languages. Smallest descriptive accounts, defined by Masuda and colleagues (Masuda & Nisbett, 2001; Senzaki et al., 2014a, 2016), were used as the coding unit. The smallest descriptive accounts were defined as the verbal account that included any meanings to describe the scene. For example, “I see two ladybugs on big green leaves” was divided into “I see two/ ladybugs/ on big/ green/leaves.” Two coding schemes were employed, with one focusing on attention to focal objects and the other focusing on social interaction among characters. These two coding schemes were not mutually exclusive; however, within each coding scheme, coding categories were mutually exclusive and exhaustive. These two coding schemes were used because previous research indicated that holistic attention involves an orientation to the visual field as a whole, including attention to relationships between focal objects and the background (Nisbett et al., 2001), and they were the primary sources of cultural differences in attention between U.S. and Japanese participants (Masuda & Nisbett, 2001). Accounts that were not related to the book (e.g., “sit down”) were excluded.

Attention to focal objects.

This coding scheme was adopted from previous research used both with adults and children in the target cultural groups (Masuda & Nisbett, 2001; Senzaki et al., 2014a, 2016). Each unique account was categorized into two groups: focal objects (including information regarding ladybugs) and the background (including information regarding anything on the page other than the ladybugs). For example, “Ten little ladybugs are eating through the leaves” were coded as three (ten/little/ladybugs/ are eating) accounts for the focal objects and one (through the leaves) account for the background. When mothers repeated the same words, they were counted as one meaningful account. For example, “The duck says quack quack” was coded as three (the duck / says / quack quack) meaningful accounts for the background object. Table 1 provides definitions and examples of categories. To examine how mothers attempted to direct their infants’ attention during shared book reading, we used the proportion of maternal narrative accounts related to the focal objects (i.e., ladybugs). If a mother only talked about the focal objects, this value would equal 100%.

Table 1.

Definitions and Examples of Maternal Narratives Coded During Shared Picture Book Reading.

Maternal narratives	Example	Definition
Focal object	“ladybugs”	Nouns
	“eating,” “sitting,” “walking”	Verbs
	“small,” “red,” “cute”	Adjectives or adverbs
Background	“butterfly,” “caterpillar,” “flowers”	Nouns
	“flying,” “moving,” “hiding”	Verbs
	“pretty,” “green”	Adjectives or adverbs
Social Interaction	“saying hello,” “good bye”	Greeting
	“the butterfly says let’s play! To the ladybugs”	Conversation
	“playing together”	Interaction

Attention to social interaction.

This coding was motivated by previous findings suggesting the social origin of cross-cultural differences in attention (e.g., San Martin et al., 2019), and the scheme was newly developed for the current study to quantify the number of times mothers discussed an interaction between main and background objects in their stories. For example, “Ladybugs say hello to the bird” and “The caterpillar is looking at the ladybugs” were qualified as social interactions, whereas “There are ladybugs and a butterfly.” was not counted as an interaction. See Table 1 for more examples.

For both coding schemes, two coders coded the responses independently. Twenty-five percent of all narratives were coded for intercoder reliability estimates, and Cohen’s kappa for the U.S. data were .89 (focal objects attention coding) and .90 (social interaction coding) and for Japanese data were .87 (focal objects attention coding) and .91 (social interaction coding). Disagreements were resolved by discussion among the coders, and the first author made the final decisions.

Results

Preliminary Analyses

As preliminary analyses, we first tested the total duration of the reading time (in seconds) and the total number of narratives (in the coding unit) across cultures. The total time spent for reading was significantly longer for Japanese dyads (M = 4 min 27 s, SD = 83.36 s) than the U.S. dyads (M = 3 min 49 s, SD = 82 s), t(101) = 2.71, p = .008. We then analyzed the total number of narrative accounts, except the maternal talk that was not related to the book (e.g., “sit down”). Unlike the total reading time, an overall number of narrative accounts was larger for the U.S. mothers (M = 151.3, SD = 59.50) than Japanese mothers (M = 115.34, SD = 45.47), t(101) = 3.46, p < .001. To control for different amounts of reading time and the total number of narrative accounts, we used the proportions of total narrative accounts in the following analysis. In the following, significant effects from omnibus tests were followed by Tukey honestly significant difference (HSD) post hoc comparisons at the p < .05 level. All t tests performed were two-tailed.

Attention to Focal Objects

We used the proportion of maternal narrative accounts related to the focal objects (i.e., ladybugs). If a mother only talked about the focal objects, this value would equal 100%. Because the dependent variable was in percentage, an arcsine transformation was conducted for analyses, although we used percentage for descriptions of means. A 2 (Culture: United States, Japan) × 2 (Gender: Boys, Girls) ANOVA with culture and gender as between-subject factors yielded a significant main effect of culture, F(1, 99) = 61.97, p < .001, ${\eta }_{\text{p}}^2=.39$. As shown in Figure 1, U.S. mothers were significantly more likely to focus on the focal objects relative to the background compared with Japanese mothers, (75.53% vs. 54.72%), t(101) = 5.65, p < .001, d = 1.17, 95% confidence interval (CI) = [0.19, 0.40]. The main effect of infants’ gender, F(1, 99) = 0.01, p = .92 nor interaction was significant, F(1, 99) = 1.64, p = .20. We also examined whether maternal socialization practices may be related to the age of infants. There were no correlations between infants’ age (in months) and maternal narrative accounts of focal objects in the United States and Japan, r = .07, p = .63, and r = −.14, p = .32, respectively. To further illustrate cross-cultural differences in maternal narratives, examples of one U.S. and one Japanese maternal narrative from the current sample are presented in the appendix.

Figure 1.

Mean percentage of maternal references to the focal objects (i.e., ladybugs) and the background across two cultural groups (i.e., 1 indicates 100% attention to the focal objects).

Note. Error bars represent ± SEM.

Attention to Social Interaction

To examine the extent to which mothers focused on social interactions among characters, we conducted a 2 (Culture: United States, Japan) × 2 (Gender: Boys, Girls) ANOVA on percentage of social interaction descriptions. There was a significant main effect of culture, F(1, 99) = 13.08, p < .001, ${\eta }_{\text{p}}^2=.12$. Japanese mothers were significantly more likely to mention social interactions among characters than the U.S. mothers (6.70% vs. 2.66%), t(101) = −3.30, p < .001, d = 0.66, 95% confidence interval (CI) = [−0.07, −0.29]. Neither the main effect of gender, F(1, 99) = 0.20, p = .66, nor interaction, F(1, 99) = 2.65, p = .11, was significant. There was also no correlation between infants’ age (in months) and maternal socialization of attention to social interaction in the United States and Japan, r = −.10, p = .54, and r = −.07, p = .60, respectively.

Sustained Attention

Infants’ sustained attention to the shared activity was measured during shared book reading, and we analyzed the percentage of infants’ looking by dividing the total looking time by the total duration of reading time. A 2 (Culture: United States, Japan) × 2 (Gender: Boys, Girls) ANOVA indicated there were no significant differences between the U.S. and Japanese infants’ sustained attention (the U.S. infants 72.12% vs. Japanese infants 69.68%), F(1, 86) = 1.09, p = .30. Table 2 presented correlations among three variables, analyzed separately for each country.

Table 2.

Correlations Among Variables for the U.S. Mother–Infant Dyads (Above the Diagonal; n = 50) and Japanese Mother–Infant Dyads (Below the Diagonal; n = 53).

	1	2	3
1. Maternal talk: Focal object	—	−.262*	−.242
2. Maternal talk: Social Interaction	−.520**	—	.244
3. Infant sustained attention	−.127	.035	—

^*p < .07 (two-tailed).

^**p < .01 (two-tailed).

Discussion

Previous research in cultural psychology has demonstrated that human cognition is a process and product of active meaning-making, and it is shaped by one’s cultural background (e.g., Bruner, 1990; Markus & Kitayama, 1991; Nisbett, 2003; Shweder, 1991). Cultural psychologists have documented a wide range of cross-cultural differences in attention in both adults and children. Developmental psychologists have also demonstrated that the sociocultural contexts impact children’s cognitive development; however, little is known about early socialization practices related to analytic–holistic patterns of attention in infancy. By using an ecologically sensitive method to examine mother–infant interactions in the United States and Japan, the current research made an important contribution to the literature. The results demonstrated that U.S. mothers were relatively more likely to direct their infants’ attention to the main objects than Japanese mothers (i.e., analytic pattern of attention). Contrarily, Japanese mothers were relatively more likely than U.S. mothers to direct their infants’ attention to the interactions between focal and background objects (i.e., a holistic pattern of attention). These findings are consistent with earlier work of cross-cultural observations between U.S. and Japanese mother–infant dyads (Bornstein et al., 1990; Caudill & Weinstein, 1969; Fernald & Morikawa, 1993; Messinger & Freedman, 1992; Morikawa et al., 1988; Shand & Kosawa, 1985; Tamis-LeMonda et al., 1992). Although these previous studies were mostly conducted in free-play settings and were not originally designed for specifically studying socialization practices of analytic–holistic patterns of attention, they provide support to the current findings by suggesting that U.S. mothers were more likely than Japanese mothers to encourage their infants to engage in object-oriented attention.

To illustrate characteristics of how the U.S. and Japanese mothers engage their 6- to 18-month-old infants in ways that foster culturally unique patterns of attention in the current study, we presented an example of maternal narratives during shared reading from each cultural group in the appendix. As illustrated in these examples, socialization contexts significantly differed for U.S. and Japanese mother–infant dyads. This finding provides a potential explanation for previous cross-cultural studies showing that young North American children tend to exhibit an analytic pattern of attention by selectively focusing on focal objects, while Japanese children tend to exhibit a holistic pattern of attention by dividing their attention to the focal and background objects, as well as attending to the relationships (e.g., Duffy et al., 2009; Imada et al., 2013; Kuwabara et al., 2011; Kuwabara & Smith, 2012, 2016; Nand et al., 2014).

The present results also found a relation between maternal references to the focal objects and social interactions. The correlation between these variables was examined separately across two cultural groups, and there was a significant negative correlation for Japanese mothers (p < .001), and a marginal trend for U.S. mothers (p < .07). In other words, mothers who talked about the background were more likely to talk about social interactions among objects in the picture book. As some researchers suggest that social orientation and socioecological differences as key factors for facilitating cross-cultural differences in attention (San Martin et al., 2019; Varnum et al., 2010), future studies should investigate the specific relation between cognitive styles and socioecological differences across cultures.

We examined maternal narratives during shared reading, because previous research has suggested that adults’ narratives created during a scene observation task correspond to the patterns of attention directly measured by an eye-tracker (Senzaki et al., 2014a), and narratives play a key role in facilitating children’s developing attention (Bruner, 1990; Wang, 2016). More recent research provides further evidence for the mediating role of socialization practices, especially involving narrative constructions, in the development of attention using neurophysiological measurements. For example, German preschool children’s attention allocation measured with an eye-tracker was related to the ways their parents verbally engaged in the scene description task with their children (Köster & Kaärtner, 2018). Our current findings provide additional support by demonstrating that maternal narratives created during shared reading, even in infancy, reflect culturally emphasized attentional patterns in two different cultural groups, and suggest that early socialization practices are the foundation of cross-cultural differences seen in children and adults. It is important for future studies to directly examine how infants and children learn how to allocate their attention through parent–child interaction. An important next step in our research is to directly observe the role of socialization in the development of infants’ cultural differences, perhaps longitudinally.

Mother–infant interaction during shared picture-book reading has been studied typically in relation to language development (e.g., Choi, 2000; Horst & Houston-prince, 2015; Ninio & Bruner, 1978; Tardif et al., 1999). The findings related to similarities and differences in language development between English- and Japanese-speaking children have been particularly important in investigating how young children learn nouns and verbs across languages (for review, see Gentner, 1982; Waxman et al., 2013). For example, U.S. mothers tended to use more noun labels, which included references to objects, animals, and people, than Japanese and Chinese mothers (Chan et al., 2009; Murase et al., 2005) during a shared reading task. On the other hand, Chinese mothers tended to use more verbs, which included references to actions made by characters, such as doing, seeing, and jumping. The results of the present study, demonstrating that U.S. and Japanese mothers directed their infants’ attention to different aspects of the book, reflect a potential source of differences in children’s language learning environments. Future research should consider the importance of examining the relation among culture, language, and cognition (Imai et al., 2016).

Only a few studies have examined cross-cultural differences in infants’ cognitive development (e.g., Azuma et al., 1981; Waxman et al., 2016). Recent research suggested that social interactions play an important role in predicting individual differences in infants’ sustained attention in the United States (e.g., Yu & Smith, 2016). Although the current study did not find significant cross-cultural differences in infants’ sustained attention, future studies should examine other cultural factors that may be linked to the development of attention and social interactions in infancy.

From a more general perspective, the current results highlight the importance of conducting cross-cultural studies with caregiver–child dyads, even in infancy. According to the sociocultural theory of child development, children learn how to think and solve problems in their everyday interactions from more knowledgeable members of their communities such as caregivers (Cole, 1996; Gauvain, 2002; Miller, 2005; Rogoff, 2003; Vygotsky, 1930/1978). Children’s development of attention provides a foundation where higher cognitive development such as vocabulary, memory, and problem solving skills develop. Although sociocultural theory emphasizes an essential role of culture in understanding human psychology and development, little attention has been given to cross-cultural investigations of parent–child interactions and how they are linked to children’s development. The present results highlight the need to conduct research from an interdisciplinary perspective. It is also important to communicate these findings with practitioners and caregivers. In the globalized world, information travels very quickly, but one type of parenting behaviors may not apply to other cultures. It is important to understand a complex and inseparable relationship between culture and child development.

One limitation of the current study was that the current sample only included mothers, especially those who were highly educated. Given the relation between culture and cognition is impacted by social class (Grossmann & Varnum, 2011; Miyamoto, 2017; Miyamoto & Ji, 2011), we recruited mothers from a similar social class for the current study. However, there are socioeconomic study (SES) differences in the amount of time parents read books to their children (e.g., Chin & Phillips, 2004); thus, it is important that future studies take a cultural psychological approach to examine the role of social class in early learning environments.

In conclusion, the present study adds to a growing body of research on cultural differences in attention by demonstrating culturally unique socialization contexts surrounding the development of infants in the United States and Japan. It has been largely assumed, without a sufficient number of empirical works, that parent–child interactions contribute to the acquisition and maintenance of cross-cultural differences in attention. The present results revealed that indeed caregiver–infant interactions were significantly different in the United States and Japan. Parents and children engage in routine activities every day, and these daily activities provide an essential environment for children to acquire culturally shared cognitive styles. Further research is needed to establish causal connections between socialization practices and cultural differences in visual attention.

Appendix

U.S. Mother Example

Page 1: Look at these little ladybugs. Should we touch them? They are all different colors—yellow, pink, and orange. They are pretty.

Page 2: There are nine ladybugs, and look at the caterpillar.

Page 3: One flew away, and now they are down to eight little ladybugs. How many orange ones are there? Do you want to touch them?

Page 4: Look, there are only 7 ladybugs left. Do you see all 7 friends? Are they trying to go home?

Page 5: 6 ladybugs are still sitting on some twigs. Look, can we count them? 1, 2, 3, 4, 5, and 6! What colors are they? Red, orange, yellow.

Japanese Mother Example

Page 1: What do you think this is? A butterfly? Yes, a butterfly says “hello!” to the ladybugs. There are many ladybugs.

Page 2: Oh here comes a caterpillar! Where do you think the caterpillar is going? What do you think these are? Yes, they are flowers. Blue and yellow flowers.

Page 3: Now we have a bird. The bird says “let’s play!”

Page 4: I think they are in the park. A cricket came in flying. The cricket said “it’s really nice out today!” and he took one ladybug to go play.

Page 5: Who will be the next? Oh here comes bees! The bees are carrying honey! Do you think they are eating the honey? Do you think bees will share the honey with ladybugs? How many ladybugs are there? 1, 2, 3, 4, 5, 6. Good job!