Grandmothers’ Developmental Expectations for Early Childhood in Botswana

Urban and rural grandmothers (n = 20) in Botswana participated in focus groups to learn their expectations for the acquisition of skills by preschool children. Their expectations for self-care, traditional politeness, and participation in household chores were dramatically earlier than developmental timetables reported for Western middle-class populations. There are some differences, however, in the urban and rural grandmothers’ expectations. Rural grandmothers had earlier expectations for self-care skills and participation in household chores, and they had more specific expectations for mastering Setswana cultural customs. In addition, some urban grandmothers, who were generally more educated, described using more reciprocal communication, and they believed in playing with their grandchildren, whereas the rural grandmothers’ communication was more instructional, and they insisted that children should play away from adults. Strikingly, there was no mention of school readiness goals or activities by either group, suggesting a “cultural misfit” between the standard early childhood curriculum, largely imported from the United States and other Western countries, and the cultural backgrounds of Batswana families. To create a more workable partnership between preschool teachers and grandparents— important caretakers of young children, both traditionally and currently—will require efforts to acknowledge and promote the values and expectations of both groups. © 2020 Wiley Periodicals, Inc. NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT, no. 170, March 2020 © 2020 Wiley Periodicals, Inc. Published online in Wiley Online Library (wileyonlinelibrary.com). ∙ DOI: 10.1002/cad.20335 93 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. 94 CROSS-CULTURAL RESEARCH ON PARENTS Childcare by grandmothers is a common feature in the traditional African milieu. In Botswana, grandmothers are perceived as custodians of culture and are implicitly expected to impart to their children and future generations traditional wisdom, values, and beliefs. Childcare by grandparents, particularly grandmothers, is widely practiced in Botswana, and in fact has historically been a core cultural practice. As Izzard (1985) explains, migration of women has been part of Botswana culture dating back to the pre-independence era, when women were heavily involved in farming and made seasonal moves between their villages and farmland. Further, some Batswana women worked in the freehold farms of Transvaal in South Africa as domestic servants. During these periods of women’s back and forth movements, grandmothers in the villages took care of children, and as Brown (1983) asserts, this arrangement was preferred because it ensured the socialization of young children in Tswana values. Nonetheless, grandmothers at that time typically played a supervisory role, as many childcare tasks were assumed by other members of the household including older children, aunts, and cousins. In contrast, Shaibu (2013) states that contemporary grandmothers in Botswana are faced with a greater challenge of supporting more children, with less help than they would have had in the past. Fathers continue to see their roles as provider, protector, and disciplinarian (Trivedi & Bose, 2018). Schatz’s (2007) study of older women’s relationships to their households in neighboring South Africa reports that on average grandmothers cared for four to five grandchildren. It is, however, surprising that although African grandmothers’ roles in childcare are widely recognized, there is very little empirical evidence on grandmothers’ beliefs and expectations regarding children’s development in this cultural context. Cross-Cultural Variability in Developmental Expectations for Early Childhood Although universal stages of childhood are recognized across cultures, parents’ expectations of the particular skills or competencies that children should have mastered at these stages vary dramatically. Recognition of such culturally constituted “developmental timetables” was an early discovery of cross-cultural research (Ninio, 1979) that has proved generative of a continuing tradition. Developmental timetables are beliefs that parents hold regarding the ages at which children should acquire and be able to use specific developmental skills (Goodnow, Cashmore, Cotton, & Knight, 1984). These timetables provide a basis for judging whether children are developing well; in Edwards, Gandini, and Giovaninni’s (1996) words, developmental timetables “measure what adults consider to be normal in the way of earliness or lateness for particular child competencies to appear” (p. 271). According to Hess, Kashiwagi, Azuma, Price, and Dickson (1980), cultural variation includes the types of competencies that adults encourage in NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT ∙ DOI: 10.1002/cad GRANDMOTHERS’ DEVELOPMENTAL EXPECTATIONS 95 children, and the level of proficiency they want children to achieve. Edwards, Gandini, and Giovaninni (1996) explain that developmental timetables are crucial for children’s development because they lead to early proficiency in culturally valued capabilities. Thus, awareness of skills expected from children across early childhood stages is important for developing supportive programs for caregivers, and for designing appropriate agendas for children in any given context. A great deal of research on developmental timetables has used the Developmental Expectations Questionnaire (DEQ), created by a team of U.S. and Japanese researchers (Hess et al., 1980). The DEQ was developed to measure adults’ expectations of child behavior in seven domains: Emotional maturity, Compliance, Politeness, Independence, School-related skills, Social skills, and Verbal assertiveness. The items, to be rated in relation to “children in general,” are to be assigned to one of three age groups (younger than 4 years, 4–6 years, and 6 years or older), thus capturing the transition from early to middle childhood. Hess et al. (1980) used the DEQ to examine maternal expectations of developmental tasks in Japan and the United States. They found that Japanese and U.S. mothers held divergent views: the mothers in Japan had earlier expectations for mastery of skills that showed self-control, compliance with adult authority, and social courtesy in interaction with adults; in contrast, U.S. mothers emphasized early acquisition of skills related to individual action, standing up for rights, and other forms of verbal assertion. Using a slightly adapted version of the DEQ, Goodnow et al. (1984) compared the developmental timetables of Lebanese-born mothers residing in Australia with Anglo-Australian mothers, and they also reanalyzed Hess et al.’s earlier data on U.S. and Japanese mothers. They found that both the Anglo-Australian mothers and the U.S. mothers had earlier expectations than did the Lebanese and Japanese mothers, particularly with regard to peer relations and verbal assertiveness. The trend toward earlier developmental expectations by parents in the United States was strongly confirmed by Edwards and colleagues’ (1996) comparative study of middle-class parents and preschool teachers in a New England town and an Italian city, with the U.S. parents expressing the very earliest expectations—especially for social and verbal development—of any study to date. In contrast, the Italian parents had later developmental expectations, and the two cultural groups of preschool teachers had expectations between the two extremes. Studies such as these support the proposition that “developmental timetables” are themselves cultural constructions expressing widely held beliefs about the nature of children’s development and parents’ roles in helping their children succeed. Group differences in developmental expectations have also been found in studies carried out in lowor middle-income countries, in which more educated, often urban mothers are compared to their less educated counterparts in rural areas, as in Williams, Jiningsih, and Williams’ (2000) report on developmental timetables in Bali. They propose that some NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT ∙ DOI: 10.1002/cad 96 CROSS-CULTURAL RESEARCH ON PARENTS low-income countries may be characterized as having “dualistic economies and social systems,” in which the more educated, affluent, and urban communities may have come to represent different cultures with regard to beliefs and practices concerning children’s development. In a similar vein, based on a study ofmothers’ developmental expectations for young children in the Philippines, Williams, Williams, Lopez, and Tayko (2000) conclude that within-culture differences in parental beliefs and expectations emanate from individual factors including parental education, family income and wealth, and access to centers of social change and modernization such as cities. Even within the same community, however—as illustrated by von der Lippe’s (1999) study of mothers’ in low-income areas of Cairo, Egypt— more educated mothers (especially those working outside the home) may have earlier developmental expectations for their children and more positive verbal interactions with them. Research on developmental timetables for young children in traditional sub-Saharan African cultural groups, using an open-ended, “emic” approach in order to explore local ideas without preset categories, indicates very early expectations for the development of responsibility. For example, Akinaware, Wilson-Oyelaran, Ladipo, Pierce, and Zeitlin (as cited in Harkness et al., 2009) carried out a survey of child care and development of young children in five different community samples in Nigeria, including rural and urban Yoruba communities, as well as other ethnic groups. Mothers in each sample were interviewed about the child’s health and behavior, maternal socialization practices, and aspirations for the child’s development. In these communities, mothers expected that by 3 years of age, children should be able to exercise control over their bodily functions, communicate clearly, and feed themselves. By about 5 years, children were considered capable of dressing themselves, running errands, and greeting elders. Such children should be able to interact outside the family in a social and helpful manner. The increased physical development of the 5-year-old, which afforded the child greater mobility, also dictated additional responsibilities: a child who was able to run down the street to play with friends would also be capable of fetching water, buying bread, and delivering messages. By 7 or 8 years, the child was deemed fully amember of society. Sevenyear-olds were expected to obey and respect their elders, care for themselves and younger siblings, and assist in the home and in the workplace. Around 8 years, children were expected to be able to think for themselves. As suggested by this research, developmental expectations for young children are embedded in larger ethnotheories of the child (Harkness & Super, 1996). An example comes from research in Kokwet, a rural community of Kenya, on mothers’ developmental expectations of when a child could be considered old enough to be sent on an errand, and when mothers thought they could know their child’s personality (Harkness & Super, 1983). The task of doing an errand, which figures prominently in research in sub-Saharan African studies (Nsamenang & Lamb, 1993; NEW DIRECTIONS FOR CHILD AND ADOLESCENT DEVELOPMENT ∙ DOI: 10.1002/cad GRANDMOTHERS’ DEVELOPMENTAL EXPECTATIONS 97 Serpell & Jere-Folotiya, 2008; Zeitlin, 1996), was generally perceived in Kokwet as a sign not only of maturity, but also a good indicator of the child’s personality. Mothers of children aged three to 10 years old were asked whether each of their children in that age range was old enough to be entrusted with a local errand and also whether the child was old enough for the mother to know “what kind of child” he or she was. For the Kokwet mothers, judgments of whether the child was old enough to go to a local shop to make a small purchase (a walk of 5–20 minutes from home) and whether the child was old enough for the mother to know its personality, increased in parallel from age 3–9 years, with about two-thirds of the mothers of 6-year-olds expressing positive judgments of both. In contrast, 70% of the New England mothers thought they could judge their child’s personality at age three, and this proportion rose to 100% for the 6-yearolds. The New England age trend in mothers’ judgments of whether the child was old enough to make a small purchase independently in one store at a local shopping mall while the mother was at a different store (an adaptation of the Kokwet measure), however, generally paralleled the Kokwet age trend but never reached 100%, even for children aged nine. Evidently, the Kokwet and Duxbury mothers were using different kinds of criteria for judging their child’s personality. For the Kokwet mothers, a child’s personality could not really be judged until the child was old enough to carry out errands responsibly: at that point, some children distinguished themselves by carrying out the errand as instructed, without getting distracted along the way. For the Duxbury mothers, in contrast, the child’s personality could be known much earlier because other aspects of behavior such as verbal communication were seen as more relevant. In summary, studies of developmental expectations in various cultures provide a window into the larger culture in which children live, while also indicating the broad parameters of what parents and others consider reasonable or normal demands of children at various ages. The present study builds on this foundation of research by focusing in particular on grandmothers in Botswana, who as both traditional guides for children’s development and increasingly as primary caretakers of their grandchildren, constitute an important but often neglected group. Specifically, we focus here on grandmothers’ developmental expectations of children between the ages of 3–5 years of age, in both rural and urban contexts. The results will be discussed as they relate to current issues in early childhood education and development in Botswana and more broadly.

C hildcare by grandmothers is a common feature in the traditional African milieu. In Botswana, grandmothers are perceived as custodians of culture and are implicitly expected to impart to their children and future generations traditional wisdom, values, and beliefs. Childcare by grandparents, particularly grandmothers, is widely practiced in Botswana, and in fact has historically been a core cultural practice. As Izzard (1985) explains, migration of women has been part of Botswana culture dating back to the pre-independence era, when women were heavily involved in farming and made seasonal moves between their villages and farmland. Further, some Batswana women worked in the freehold farms of Transvaal in South Africa as domestic servants. During these periods of women' s back and forth movements, grandmothers in the villages took care of children, and as Brown (1983) asserts, this arrangement was preferred because it ensured the socialization of young children in Tswana values. Nonetheless, grandmothers at that time typically played a supervisory role, as many childcare tasks were assumed by other members of the household including older children, aunts, and cousins.
In contrast, Shaibu (2013) states that contemporary grandmothers in Botswana are faced with a greater challenge of supporting more children, with less help than they would have had in the past. Fathers continue to see their roles as provider, protector, and disciplinarian (Trivedi & Bose, 2018). Schatz' s (2007) study of older women' s relationships to their households in neighboring South Africa reports that on average grandmothers cared for four to five grandchildren. It is, however, surprising that although African grandmothers' roles in childcare are widely recognized, there is very little empirical evidence on grandmothers' beliefs and expectations regarding children' s development in this cultural context.

Cross-Cultural Variability in Developmental Expectations for Early Childhood
Although universal stages of childhood are recognized across cultures, parents' expectations of the particular skills or competencies that children should have mastered at these stages vary dramatically. Recognition of such culturally constituted "developmental timetables" was an early discovery of cross-cultural research (Ninio, 1979) that has proved generative of a continuing tradition. Developmental timetables are beliefs that parents hold regarding the ages at which children should acquire and be able to use specific developmental skills (Goodnow, Cashmore, Cotton, & Knight, 1984). These timetables provide a basis for judging whether children are developing well; in Edwards, Gandini, and Giovaninni' s (1996) words, developmental timetables "measure what adults consider to be normal in the way of earliness or lateness for particular child competencies to appear" (p. 271). According to Hess, Kashiwagi, Azuma, Price, and Dickson (1980), cultural variation includes the types of competencies that adults encourage in children, and the level of proficiency they want children to achieve. Edwards, Gandini, and Giovaninni (1996) explain that developmental timetables are crucial for children' s development because they lead to early proficiency in culturally valued capabilities. Thus, awareness of skills expected from children across early childhood stages is important for developing supportive programs for caregivers, and for designing appropriate agendas for children in any given context.
A great deal of research on developmental timetables has used the Developmental Expectations Questionnaire (DEQ), created by a team of U.S. and Japanese researchers (Hess et al., 1980). The DEQ was developed to measure adults' expectations of child behavior in seven domains: Emotional maturity, Compliance, Politeness, Independence, School-related skills, Social skills, and Verbal assertiveness. The items, to be rated in relation to "children in general," are to be assigned to one of three age groups (younger than 4 years, 4-6 years, and 6 years or older), thus capturing the transition from early to middle childhood. Hess et al. (1980) used the DEQ to examine maternal expectations of developmental tasks in Japan and the United States. They found that Japanese and U.S. mothers held divergent views: the mothers in Japan had earlier expectations for mastery of skills that showed self-control, compliance with adult authority, and social courtesy in interaction with adults; in contrast, U.S. mothers emphasized early acquisition of skills related to individual action, standing up for rights, and other forms of verbal assertion. Using a slightly adapted version of the DEQ, Goodnow et al. (1984) compared the developmental timetables of Lebanese-born mothers residing in Australia with Anglo-Australian mothers, and they also reanalyzed Hess et al.' s earlier data on U.S. and Japanese mothers. They found that both the Anglo-Australian mothers and the U.S. mothers had earlier expectations than did the Lebanese and Japanese mothers, particularly with regard to peer relations and verbal assertiveness. The trend toward earlier developmental expectations by parents in the United States was strongly confirmed by Edwards and colleagues' (1996) comparative study of middle-class parents and preschool teachers in a New England town and an Italian city, with the U.S. parents expressing the very earliest expectations-especially for social and verbal development-of any study to date. In contrast, the Italian parents had later developmental expectations, and the two cultural groups of preschool teachers had expectations between the two extremes. Studies such as these support the proposition that "developmental timetables" are themselves cultural constructions expressing widely held beliefs about the nature of children' s development and parents' roles in helping their children succeed.
Group differences in developmental expectations have also been found in studies carried out in low-or middle-income countries, in which more educated, often urban mothers are compared to their less educated counterparts in rural areas, as in Williams, Jiningsih, and Williams' (2000) report on developmental timetables in Bali. They propose that some low-income countries may be characterized as having "dualistic economies and social systems," in which the more educated, affluent, and urban communities may have come to represent different cultures with regard to beliefs and practices concerning children' s development. In a similar vein, based on a study of mothers' developmental expectations for young children in the Philippines, Williams, Williams, Lopez, and Tayko (2000) conclude that within-culture differences in parental beliefs and expectations emanate from individual factors including parental education, family income and wealth, and access to centers of social change and modernization such as cities. Even within the same community, however-as illustrated by von der Lippe' s (1999) study of mothers' in low-income areas of Cairo, Egyptmore educated mothers (especially those working outside the home) may have earlier developmental expectations for their children and more positive verbal interactions with them.
Research on developmental timetables for young children in traditional sub-Saharan African cultural groups, using an open-ended, "emic" approach in order to explore local ideas without preset categories, indicates very early expectations for the development of responsibility. For example, Akinaware, Wilson-Oyelaran, Ladipo, Pierce, and Zeitlin (as cited in Harkness et al., 2009) carried out a survey of child care and development of young children in five different community samples in Nigeria, including rural and urban Yoruba communities, as well as other ethnic groups. Mothers in each sample were interviewed about the child' s health and behavior, maternal socialization practices, and aspirations for the child' s development. In these communities, mothers expected that by 3 years of age, children should be able to exercise control over their bodily functions, communicate clearly, and feed themselves. By about 5 years, children were considered capable of dressing themselves, running errands, and greeting elders. Such children should be able to interact outside the family in a social and helpful manner. The increased physical development of the 5-year-old, which afforded the child greater mobility, also dictated additional responsibilities: a child who was able to run down the street to play with friends would also be capable of fetching water, buying bread, and delivering messages. By 7 or 8 years, the child was deemed fully a member of society. Sevenyear-olds were expected to obey and respect their elders, care for themselves and younger siblings, and assist in the home and in the workplace. Around 8 years, children were expected to be able to think for themselves.
As suggested by this research, developmental expectations for young children are embedded in larger ethnotheories of the child (Harkness & Super, 1996). An example comes from research in Kokwet, a rural community of Kenya, on mothers' developmental expectations of when a child could be considered old enough to be sent on an errand, and when mothers thought they could know their child' s personality (Harkness & Super, 1983). The task of doing an errand, which figures prominently in research in sub-Saharan African studies (Nsamenang & Lamb, 1993;Serpell & Jere-Folotiya, 2008;Zeitlin, 1996), was generally perceived in Kokwet as a sign not only of maturity, but also a good indicator of the child' s personality. Mothers of children aged three to 10 years old were asked whether each of their children in that age range was old enough to be entrusted with a local errand and also whether the child was old enough for the mother to know "what kind of child" he or she was. For the Kokwet mothers, judgments of whether the child was old enough to go to a local shop to make a small purchase (a walk of 5-20 minutes from home) and whether the child was old enough for the mother to know its personality, increased in parallel from age 3-9 years, with about two-thirds of the mothers of 6-year-olds expressing positive judgments of both. In contrast, 70% of the New England mothers thought they could judge their child' s personality at age three, and this proportion rose to 100% for the 6-yearolds. The New England age trend in mothers' judgments of whether the child was old enough to make a small purchase independently in one store at a local shopping mall while the mother was at a different store (an adaptation of the Kokwet measure), however, generally paralleled the Kokwet age trend but never reached 100%, even for children aged nine. Evidently, the Kokwet and Duxbury mothers were using different kinds of criteria for judging their child' s personality. For the Kokwet mothers, a child' s personality could not really be judged until the child was old enough to carry out errands responsibly: at that point, some children distinguished themselves by carrying out the errand as instructed, without getting distracted along the way. For the Duxbury mothers, in contrast, the child' s personality could be known much earlier because other aspects of behavior such as verbal communication were seen as more relevant.
In summary, studies of developmental expectations in various cultures provide a window into the larger culture in which children live, while also indicating the broad parameters of what parents and others consider reasonable or normal demands of children at various ages. The present study builds on this foundation of research by focusing in particular on grandmothers in Botswana, who as both traditional guides for children' s development and increasingly as primary caretakers of their grandchildren, constitute an important but often neglected group. Specifically, we focus here on grandmothers' developmental expectations of children between the ages of 3-5 years of age, in both rural and urban contexts. The results will be discussed as they relate to current issues in early childhood education and development in Botswana and more broadly.

Methods
Research Setting. This study was conducted in two different locations in Botswana: Gaborone and Kanye.
Gaborone. The city of Gaborone was established at the time of independence in 1966. Gaborone is positioned in the flat valley between Kgale and Oodi hills, on the west of Notwane River, about 15 kilometers from the South African border (Mosha, 1996;Sebego & Gwebu, 2013). On the periphery of the city are eight villages that serve as residential suburbs for a significant proportion of people who work in Gaborone (Sebego & Gwebu, 2013). From an administrative center with a few landmarks including a small railway station in 1962, Gaborone has become one of the fastest growing capitals in Africa. Mosha (1996) attributes the city' s growth and sustained development to its careful planning and management. Gaborone has modern civic and commercial centers, infrastructure such as water, electricity, roads, and sewage system (Mosha, 1996), as well as clinics, hospitals, and schools. In addition to having several public and private universities, numerous public and private high schools, and a multitude of elementary schools, Gaborone has about 107 licensed preschools.
Along with administrative and infrastructure developments, Gaborone has experienced an upsurge in population since independence, from 3,855 in 1964 (Sebego & Gwebu, 2013) to 23,192 in 2011(Government of Botswana, 2015. According to the 2011 census, Gaborone has about 120 neighborhoods, arranged following the Botswana Social Integration Policy, which aims to plan localities in a systematic manner that is free from polarization of social classes, races, and income groups (Mosha, 1996). Thus, neighborhoods contain a mixture of plots/houses for different income groups, which are allocated or sold to individuals of different social classes and ethnicities, based on availability. The constitution of neighborhoods in the city of Gaborone thus differs greatly from the living arrangement in traditional communities like Kanye.
Kanye. Kanye is a village of the Bangwaketse tribe and is located about 53 miles southwest of Gaborone. It is one of the largest villages in Botswana and is the headquarters of the southern district (Mbata, 2006). The village was established in the 1790s by Kgosi (Paramount Chief) Makaba. Kanye lies on a series of hills that stretch to the bushveld of South Africa. According to Botswana' s 2011 population census, Kanye' s population was 52,214. Although Kanye' s population is greater than that of some cities, it is called a village because of its tribal administration. The village is led by a paramount chief, Kgosikgolo Malope Gaseitsewe II. Kanye is organized by wards. The original wards are made of occupants who share relations and are led by headmen (subordinates of the paramount chief), while newly established wards are made up of non-related individuals and families. The royal ward kgosing is located at the top of the hill. Among other facilities and services, Kanye has modern infrastructure including water, electricity, roads, a mission hospital, several clinics, banks, a technical college, a public high school, several public and private elementary schools, and about 31 licensed preschools. Kanye also hosts the country' s National Food Technology Research Center.
Participants. Participants were 20 grandmothers of preschool-going children whose ages ranged between 3 and 5 years. All the grandmothers who participated in this study were at least 50 years of age or older. Ten grandmothers were selected from Gaborone city and ten were recruited from Kanye. Rural grandmothers were generally older and mostly not employed outside the home, in contrast to urban grandmothers (Table 5.1). These grandmothers were mostly the primary caregivers for their grandchildren, whose mothers were working in towns and cities away from their villages. A few of the rural grandmothers reported being retired while the majority stated that they never worked outside the home. Other than growing their own food and relying on their pension, the rural grandmothers mentioned that their own children served as the grandmothers' principal support. In contrast, the urban grandmothers were more educated, were mostly employed on a full-time basis, and some were co-resident with their adult children. These grandmothers had both rural and urban experiences from spending their childhood years in their respective villages and later residing in Gaborone city for purposes of higher education and employment. All the participating grandmothers were recruited through a snowball process whereby the first grandmothers were referred to the study by parents and preschool teachers who knew about the study. Others were referred by the grandmothers who had agreed to participate in the study. The researcher used the referred contact information to follow up with potential participants.

Measures.
A protocol developed for this study was used to guide grandmothers' focus group discussions (Krueger & Casey, 2015). The protocol consisted of nine open-ended questions that included asking grandmothers about the skills they expected children to have gained when they are three to 5 years old, why the grandmothers considered these skills important, and how they promoted them. In addition, a demographic background questionnaire was used to gain general information about the grandmothers, including their age, occupational status, educational background, and the age of their preschool-attending grandchild.
Procedures. The first author collected data for this study between July and August 2018, after obtaining approvals from the University of Connecticut Institutional Review Board and from the Ministry of Local Government in Botswana. A focus group discussion for Gaborone grandmothers was conducted at a seminar room located within the University of Botswana. For Kanye grandmothers, the focus group discussion was held at a rented hall located in Kanye village. Each of the participating grandmothers completed the consent form and then the demographic questionnaire in the presence of the researcher. Grandmothers had a choice of completing the original, English questionnaire or one translated into Setswana, the local language. The translated Setswana focus group guide and demographic questionnaire were used for participants who could not read and/or write English language, and for all participants who preferred to use a translated instrument. The researchers and research assistants helped participants who could not read and/or write to complete the consent form, and then the questionnaire, by reading aloud each item to the participants, waiting for their response, and checking the item verbally selected by the participant. The Kanye focus group discussion was conducted in Setswana, while both Setswana and English were used during the Gaborone focus group discussion. Each focus group discussion lasted for 45-60 minutes. Discussions were audio recorded with participants' permission. Participants were served with refreshments, and each participating grandmother was given a $5 equivalent amount as compensation for their transport cost.
Data Analysis. Focus group data were analyzed inductively following the steps proposed by Cresswell (2014), which include organizing, preparing, and transcribing the data verbatim; reading through the selected scripts of data to get a general sense of the information; and reflecting on its meaning. Coding of data followed thematic analysis (McClelland, 1975) and open coding (Strauss & Corbin, 1990). Each transcription was independently analyzed and checked for accuracy by the primary investigator. A three-step approach to thematic analysis was used which involved: (a) Reading the entire text to make sense of the data, (b) looking for statements or phrases to sort the data into segments, and (c) line-by-line reading and assigning codes to segments of text. The coding process included extracting and assigning codes to themes and sub-themes. Overlapping themes were combined and checked for accuracy by the research team. For quality and validity, the author discussed the coding process with the main supervisor. The first and second authors then reviewed each code and theme to reach a consensus, and finally, verified that themes were supported by codes and quotations. The coding process was done on the focus group transcripts in their original language version. The extracted quotations were then translated to English language for reporting purposes. The demographic questionnaire responses were summarized in terms of means and range of response for each item, separately for the urban and rural groups.

Results
The grandmothers described an assortment of skills that they expected their grandchildren to demonstrate at the ages from 3 to 5 years. Grandmothers in both groups explained that while they have general expectations for children at different ages, they were also mindful of the fact that some children show the expected skills earlier, while others develop the skills much later, which they attributed to individual differences. As an urban grandmother explained: "Before you get to that, let me say that for some things we will be saying we expect this from a child at 3 years, but children are different. There are some children who demonstrate these skills before three years, showing the skills at one year… just that children are different." Five recurring themes in grandmothers' expectations for their grandchildren were identified: self-care skills, communication skills, learning Setswana customs, starting to do household chores, and environmental independence. Four out of the five identified themes for this study were similar to some of the categories of the Developmental Expectations Questionnaire (Hess et al., 1980)-self-care, compliance, communication, and environmental independence. Grandmothers in both groups further differentiated between skills they expected children to attain by age 3, and those that were expected for children aged 4-5 years.
Expected Skills at 3 Years. Self-Care. Grandmothers in both groups emphasized that they expected their 3-year-old grandchildren to be able to feed themselves, to be potty-trained, and to be at the stage where they start dressing and undressing themselves. A rural grandmother commented, "They should be able to talk, and feed themselves while you are observing that they are eating well and they get satisfied." Similar sentiments were expressed by an urban grandmother: "They should be able to feed and dress themselves at three years. At three years they have already started because a child can remove their own panty and also put it on." A rural grandmother said, "Even at two years, this one here [pointing to a nearby child], he is two years going for three. He knows how to dress himself and doesn't wear clothes backwards; he knows how to put on a trouser, socks and shoes. Yes, he dresses himself." Regarding potty-training, an urban grandmother explained, "They should be able to undress and go to the toilet," and a grandmother in the rural focus group stated, "When they are small, when we take them out of confinement, they leave the diapers altogether." Confinement, or botsetsi as it is popularly known in Botswana, is a practice wherein the new mother and infant are secluded for a period of 6-12 weeks (Merriweather, 1992) for two main purposes: (a) To strengthen and protect the mother and the infant, and (b) to grant the new mother and their infant time to bond (Geiger & Alans, 2005). The rural grandmothers thus expected their grandchildren to be off the diapers from as early as 1.5 months.
Communication Skills. Both rural and urban grandmothers explained that they expected their 3-year-old grandchildren to be able to talk and to follow simple commands. One rural grandmother said, "They should be able to talk and take porridge and eat." An urban grandmother explained, "He can be sent around; in Setswana culture a child starts to be sent to do tasks at three years. We send them around: 'Go and get me that thing, bring me that thing.'" A similar comment was made by a rural grandmother, "He is going on three, but he is clever. This friend of mine, you see him, I send him to the house, 'Go and get me something from the house.'" A rural grandmother described even earlier competence in communication: "This one is two years, …he is clever, and he is the one we send around because there are no other people. Yes, he already understands everything…." The word "clever" used by these grandmothers (botlhale) is closely linked with social competence reflecting the child' s ability to perform household tasks and other valued skills, such as giving a visitor a chair. At no given time did the grandmothers from either group use the term in reference to academic intelligence. This is line with the literature on African concepts of child competence. ''Socialization is not organized to train children for academic pursuits," write Nsamenang and Lamb (1994, p. 137), "or to become individuals outside the ancestral culture. Rather, it is organized to teach social competence and shared responsibility within the family system and the ethnic community." This concept of intelligence departs from models of intelligence widely found in Western cultures, but is similar to the terms cejela and tumikila found in Zambia (Serpell, 1993), and ng'om among the Kipsigis of Kenya (Super & Harkness, 1986). It also relates to the nationally valued concept of botho in that it recognizes the rights and the responsibilities of all people (Dube et al., 2016), including young children. The linkage of botlhale and botho is further elucidated in the Botswana Vision 2016 (Government of Botswana, 1996) which defines botho as "process of earning respect by first giving it, and of gaining empowerment by empowering others." Thus, being "clever" as characterized by the grandmothers here includes socialization into respecting others and shows nurturance of ideals that promote botho.
Learning Setswana Customs. Taking with both hands when given something and being polite were recurring sub-themes during focus group discussions of both urban and rural grandmothers. Grandmothers stressed that they strive to impart cultural customs to their grandchildren from an early age. An opinion by an urban grandmother was, "In Setswana culture we have different values that we want to start instilling in children when they are three years, like the culture of respect-in the form of ma'am, sir, and the like." A rural grandmother remarked: "Yes, you tell them. When you give the child and they receive, take with both hands, and say 'Thank you, Mama.'" The rural grandmothers expanded the list of Setswana customs mentioned by the urban grandmothers, adding that they want their 3-yearold grandchildren to be able to offer a chair to a visitor, to know right from wrong, and to know that they should play away from adults, citing the latter as a sign of respect for the elders. A rural grandmother explained: "The first thing when they have become clever, you teach them that when visitors arrive at home, you get them a chair, visitors should not come and stand like this [demonstrating]." Another commented on playing away from adults: "From two years…, one year, you give them toys to play there…, you watch them from a distance. You watch them, yes, but they should not be mixing with adults, yes. If you allow them to play near you, they will get used to it." In contrast, an urban grandmother commented, "We grandmothers play with our grandchildren, they come to you wanting to play and you put everything aside and play with them. Or you take the child and put them on your lap and talk to them and I think that builds the connection." Beginning Household Chores. The two groups of grandmothers were adamant that they expected their grandchildren to start participating in household chores at about 3 years of age. In particular, the grandmothers expected their grandchildren to be able to wash their socks and panties, and to be able to take their dishes to the sink after eating. An urban grandmother said, "When washing clothes, you put their sock in the laundry, allow the child to wash the sock, while knowing that you will finish it off, even their panty if it' s a girl, you allow them to wash knowing that you will eventually rinse it properly. But the child would know that they should wash their panty." A rural grandmother commented: "You see this one, I started with her when she started going to crèche. I started with socks; when she comes from school, I put socks for her and tell her, 'Put them in a bath.'" Regarding taking the plate to the sink, an urban grandmother said: "When they finish with the dish, they take it to the kitchen." A rural grandmother extended this expectation a bit further: "I just started teaching her today that a dish is not left in the sink." She says, 'Should I wash it mama?' And I say 'Yes.' Even if she does not clean it properly, I will rinse it to make it clean." Environmental Independence. In addition to the above, the rural grandmothers also discussed environmental independence and children' s safety. The discussion yielded divided views regarding this issue. There were some rural grandmothers who expected children to start playing outside the yard, and going to buy from a neighborhood tuck-shop (a small grocery store located within the neighborhood) unaccompanied by adults from 3 years or earlier. As one of the grandmothers remarked, "For me at two years the child goes to the tuck-shop, because they can talk, yes I send them, 'Go and buy bread.'" Another said: "I don't send him, when he is given money he goes to the tuck-shop and buys whatever his heart wants." Other rural grandmothers expressed discomfort with the idea, stating that environmental independence for young children was an expected skill in the past, but that it is no longer safe because of child abduction and vehicles that speed through neighborhoods. In expressing her difference of opinion about this issue, a rural grandmother said: "For me, at my house, there is a tuck-shop just here [pointing with a finger], in front…, my grandchildren don't go out, I fasten the gate, yes!" Expected Skills at 4-5 Years.
Self-Care. The grandmothers from both locations made it clear that they expected their grandchildren to demonstrate higher skills in selfcare at 4 and 5 years. According to both groups of grandmothers, children should be able to bathe themselves when they are 4-5 years: "Even bathing themselves, they bathe themselves at four to five years, and your role is to help them with their back, but as for the face and legs, even to apply body lotion…, they know that it is done and even how to do it," as an urban grandmother explained. A rural grandmother summarized all her expectations as, "When they are five years old, they know how to feed themselves, they have accomplished to talk, they have accomplished bathing themselves, and have accomplished knowing right from wrong." Household Chores. Urban grandmothers expected their grandchildren to demonstrate advanced skills in household chores. As one specified, "When a child is five years we expect them to be able to make their bed, be able to fold their clothes and put them away nicely." Another expectation that urban grandmothers had for their grandchildren at 5 years was to be able to participate in household cooking. One grandmother said: "They will ask you, can I peel the potatoes? You would say 'Peel …' You give them and tell them that, 'You peel it this way.'" The rural grandmothers had even higher expectations for their 5-yearold grandchildren: they expected them to be able to make tea and to begin to cook. As one grandmother said, "They know that when visitors come, they make tea, they prepare cups and bring tea." Another commented: "Yes, right now even soft porridge, they know how prepare it." Environmental Independence. As with their concerns about 3-year-old grandchildren, the rural grandmothers insisted that they were prevented from allowing their children to play outside in contemporary times because of safety concerns. One grandmother had this to say: "Nowadays it is dif-ficult… you know, to open [the gate] for a child to go and play outside is not good. Yes, unless they sneak out, but if I glance back and don't see them, I open the gate and call them back." To sum up the reasons why these grandmothers were uncomfortable with children' s environmental independence, one grandmother explained: "…cars, child abduction, children get lost…" If it were not for safety concerns, the rural grandmothers made clear, "When a child is 5, they could go and play with others outside the yard." In general, urban and rural grandmothers expected similar developmental skills from their grandchildren. However, the timing of the expected skills was comparatively earlier for rural than for urban grandmothers.
Further, the rural grandmothers expected higher levels of competencies in household chores for their grandchildren at the ages of 4 and 5 years than the urban grandmothers.
Grandmothers' Views on the Importance of the Expected Skills. Regarding why they deemed the expected skills important, grandmothers from urban and rural locations had some similar but also some different views. The main themes that came out of the focus group discussions were independence, good behavior, and knowledge of who they are.
Independence. Both groups of grandmothers felt that the skills they expect from their grandchildren are crucial for the development of children' s independence. An urban grandmother explained: "In Setswana culture, independence is a very important skill, what we really want…, what we teach them. … by involving them to do the things by themselves so that they are not dependent on you all the time: 'Come and dress me, come and do this.'" Relatedly, a rural grandmother had this to say: "Yes, we encourage these skills so that the children can use them later in life as they grow, knowing that when they are given food they should first wash their hands, then receive food using both hands, eat, and wash hands, then drink water, so that they don't stay hungry or thirsty when I am not there." Another rural grandmother added, "The children should be able to talk for them-selves…, when they are left with other people." From these grandmothers' point of view, the expected skills provide their grandchildren with the tools essential for survival and for becoming self-reliant as adults. Indeed, the theme of independence was a crosscutting topic during the urban and rural grandmothers' focus group discussions. The discussions of the expectations that were described earlier-self-care, communication, Setswana customs, participation in household chores and environmental independence-all had an implicit theme of independence.
Good Behavior. Although the urban grandmothers did not discuss "good behavior" during the deliberation on why they considered the skills they expect from their grandchildren to be important, there was no doubt that they endorsed and promoted good behavior by their grandchildren. These grandmothers expressed their enthusiasm for good behavior when they talked about Setswana customs. For rural grandmothers, good behavior was expressed as one of the core values of Setswana culture. According to these grandmothers, children express good behavior in several ways: "Good behavior…when you are with the child, they have to show good manners. When you come as visitors, and my grandchild is home, they are not supposed to be playing in your midst. They know: No! I should have taught them to exercise restraint, to be obedient." Know Who They Are. This theme did not come up during the discussion with urban grandmothers, but was discussed by rural grandmothers. The rural grandmothers emphasized that it is important for their grandchildren to know who they are. That way they can be able to identify themselves, their parents by names, and where they come from: "A child this size [showing size with a hand], when you ask them, 'Who are your parents?' They should not say, 'Mama, Papa' that is not good! You should teach the child that they are a child of Dineo-your real name, not 'Mama, Papa.' When I ask, 'Whose child are you?' you should be able to say, 'I am the child of Tiso from such and such a place.'" Grandmothers' Practices to Promote Acquisition of Children's Skills. The grandmothers described a host of practices that they used to promote the skills they expected from their grandchildren. In addition, both urban and rural grandmothers mentioned love, affection, patience, persistence, presence, and attention as some of the characteristics that draw children closer to their grandmothers. As one urban grandmother commented, "The patience, the affection, boundless love attracts children to grandmothers because the mothers are too busy. Even when the mother arrives home from work, it' s 'Hi,' then she goes to the bedroom and at times she would stay in the bedroom without giving the child attention… and grandmother would sit and play with the child." A rural grandmother echoed a similar sentiment: "To raise a child requires someone who is patient, and who has lots of love for the child, you will see the child growing well and also loving you." Both urban and rural grandmothers supported the development of desired skills among their grandchildren by teaching or training, guiding, demonstrating, modeling, and participating. A rural grandmother stated, "When a child starts learning how to talk, we begin teaching them many things, like, 'Ma'am,' 'Sir,' 'Come here, go there.'" Another rural grandmother added: "Every time after the child bathes, you show them, how to put on a t-shirt, panty, trouser, shoes, they get used to it. Even shoes, you guide them how to put on shoes properly, when they mix them, you tell them 'Hold the shoes this way and put this one this side.'" In explaining a similar approach, an urban grandmother stated, "When we talk to children, we show them that this goes this way, this that way, the head goes here, the arm goes here, so that they can know what to do the next time they are on their own." Another skill-developing practice used by the grandmothers was detailed by an urban grandmother: "I believe there should be a child' s role model in the home, one who can model the behaviors that the child would believe in. And it is very important that we model the behavior and also be a positive role model." Similarly, a rural grandmother commented, "Yes, appropriate role model because if you insult, saying 'You such and such,' even the child will say it thinking that is the language." The grandmothers in both urban and rural settings were in agreement that teaching the children expected skills can be achieved by involving and allowing the children to engage and learn through participating. Nonetheless, the two groups of grandmothers differed on when to start training their grandchildren. Urban grandmothers' timing for verbal communication and interaction with their grandchildren was earlier than that of rural grandmothers. In contrast, rural grandmothers had earlier expectations for self-help skills like potty-training, dressing self, cooking, and running errands. While urban grandmothers explained their expected age for dressing self as "They dress themselves at three years," a rural grandmothers made it clear that: "Even at two years…, he knows how to dress himself and doesn't wear clothes backwards." This variation in timing of expected skills influenced grandmothers' ideas about when they should start helping the children to attain such skills.
Potty-training was another area where there were clear differences of timetable between the two groups of grandmothers. Descriptions given by rural grandmothers included "When we take them out of confinement, they leave diapers altogether," and "For me, all my grandchildren, when they are still in confinement, the moment the umbilical cord detaches, I train them to squat. I don't have a child… altogether, I don't have a child who defecates on clothes. I start training them from a month." The remarks made by these rural grandmothers illustrate the earliness of their pottytraining, which contrasts with potty-training timing by the urban grandmothers, who expected their grandchildren to "be able to undress and go to the toilet" at 3 years. Another difference in expectations between the two groups of grandmothers regards participation of children in cooking. Rural grandmothers expected their grandchild to be able to make tea and soft porridge at the age of 5 years-quite an advanced skill-whereas urban grandmothers expected the child to be able to at least peel vegetables for adults who were cooking. Running errands was expected as early as 2 years by rural grandmothers: "For me at two years the child goes to the tuck-shop, because they can talk, yes I send them, go and buy bread." In contrast, urban grandmothers started a bit later: "A child starts to be sent to do tasks at three years." There were some differences between the two groups of grandmothers in how they communicated with their grandchildren. An urban grandmother said, "The grandmother would rather say, 'My boy, wait, let' s sit down and talk,' …. You see, sometimes children want you to reason with them." Another urban grandmother suggested offering praise: "After the child has been able to put on a garment well, 'Wow my beautiful smart girl!' You see, motivating her." This grandmother further elaborated, "When I come from work, I have time for them, I sit down with them and we play…, and do this and that. Another said, "You put things aside, take the child and put her on your lap and talk to her." These descriptions of verbal communication given by urban grandmothers represent reciprocity in communication, warmth and flexibility, negotiation and bargaining, as well as emotional closeness. For rural grandmothers, communication appeared more instructional, often in the form of commands: "You tell them, 'Take with both hands, and say thank you.'" Another rural grandmother elaborated: "At one year, you give them toys to play there…, you watch them from a distance." The same sentiment was emphasized by another rural grandmother, "You have to teach them that when we are here, they play there!" As described in these grandmothers' expressions, their way of communication required compliance and very little verbal response from their grandchildren. The rural grandmothers' insistence on children playing away from adults contrasts with the urban grandmothers' ideas of playing with grandchildren, a form of proximal interaction.

Discussion and Conclusion
Both urban and rural grandmothers in the present study described an assortment of expected skills for children at the ages of 3-5 years, with interesting differences in their apparent expectations. From a cross-cultural perspective that the Batswana grandmothers' expectations for the acquisition of skills related to self-care, traditional Setswana rules related to politeness, and participation in household chores were dramatically earlier than developmental timetables seen in research with Western middle-class populations (Joshi & MacLean, 1997;Winskel, Salehuddin, & Stanbury, 2013). Interestingly, even the domain of communication was described in terms of the child' s ability to listen, understand, and carry out a requested errand, and a child who could do this even before the age of 3 years was described as "clever" (botlhale). This finding of rural grandmothers' communication with their grandchildren corroborates an earlier report by Geiger and Alans (2005), which suggests that verbal communication between adult caregivers and children in Botswana traditionally was largely instructional, with very little verbal response encouraged or expected from the child. The style of communication described by rural grandmothers also aligns with the Botswana nation' s value of botho or integrity, humanity and dignity for all, which is nurtured through training children to listen and to take instructions as expressed by adults without question. This present finding is consistent with the results of Zeitlin' s research in Nigeria, referenced above, in which the errand figures largely as a tool of socialization for children; it is also similar to Harkness and Super' s (1983) research, also cited above, on the importance of the errand as an indicator of the child' s personality in a rural Kenyan village. Similarly, Harkness and Super' s (1977) report on child language acquisition in the same village concluded that comprehension, not production, was the primary goal of language socialization practices.
Given the method of sampling and the exclusive use of focus groups for data collection, caution is required in generalizing these results to actual behavior, and to other samples. Nevertheless, based on such similar findings from three different communities of West, South, and East Africa, we might speculate that such developmental expectations are widely shared across the various cultural groups in sub-Saharan Africa. Our study reveals both similarities and differences, however, in the types of skills that urban and rural grandmothers expected from their grandchildren at different ages.
With regard to expected skills at 3 years, the grandmothers studied here, from urban and rural locations, generally expected the same skills. The rural grandmothers, however, appeared to have earlier expectations for self-care skills and participation in household chores, and they elaborated extensively in describing their expectations for mastering Setswana cultural customs. Notably, grandmothers in both settings explained that the main purpose of assigning household chores to such young children was for training purposes rather than for actual help: if the child did not properly rinse her panties after washing them, or failed to wash her plate completely, the grandmother would quietly take care of it for her.
For interacting with their grandchildren, some urban grandmothers described using more reciprocal communication, and they believed in playing with their grandchildren, whereas the rural grandmothers thought of their communication with their grandchildren as more instructional, and they insisted that children should play away from adults. The urban grandmothers who talked about playing with their grandchildren were more educated. A striking absence in both the rural and urban grandmothers' focus group discussions was any mention of school readiness goals or activities, such as book reading with their grandchild. In contrast to the current preoccupation with getting children ready for school as young as 2 years of age in the United States, school readiness was apparently not on the developmental agenda of these grandmothers, even though they all had grandchildren attending preschool.
Formal early childhood development (ECD) programs are a new addition to the national educational system in Botswana, and they lack many of the resources that are standard in post-industrial societies, including ageappropriate educational and play materials, furniture, trained ECD teachers, and even adequate building spaces. These issues are well recognized, with both local and international efforts to address them. Less well recognized, however is the "cultural misfit" between the standard ECD curriculum, which is largely imported from the United States and other Western countries, and the cultural backgrounds of the families who send their children to preschool. A few observers have highlighted this issue, suggesting that a refocusing of curricula on local expectations may yield valuable results Ng'asike, 2014;Serpell & Nsamenang, 2014). The issue is not only the expected age for particular accomplishments, but also more profoundly the selection of behaviors of most interest. Thus, even excellent psychometric examination of developmental assessment tools originating in Western industrial cultures (e.g., Hsiao et al., 2017) may fail to recognize the developmental agenda of families.  have highlighted as well the ethical issues of exporting educational interventions for parents.
As we have shown in the present paper, Batswana grandmotherswho are often the primary or shared caregivers of their grandchildrenhave a very different developmental agenda in mind for young children, even for those who are already enrolled in preschool. From comments made by the grandmothers in the focus groups, it appears that preschool teachers were at best keeping them informed about their grandchild' s progress, but they apparently did not offer advice on how the children could also learn school-related skills at home. Parents and grandparents in Botswana may feel that sending their young child to preschool entails making a painful choice between inculcating behavioral norms and skills that have sustained their culture for centuries, and preparing the child for an unknown future in a world that may still seem strange. For these caregivers, and others like them in a variety of settings, a workable partnership between families and schools will require efforts to acknowledge, validate, and implement socialization practices that capture the best of both worlds.