1. Art and Diseño

IMHJ (Wiley)
RIGHT
BATCH
top of AH
A
R
T
I
C
L
E
ATTACHMENT, CULTURE, AND THE CAREGIVING
SYSTEM: THE CULTURAL PATTERNING OF
EVERYDAY EXPERIENCES AMONG ANGLO AND
PUERTO RICAN MOTHER– INFANT PAIRS
VIVIAN J. CARLSON AND ROBIN L. HARWOOD
University of Connecticut
ABSTRACT: This investigation focuses on cultural differences in the relationship between maternal sensitivity, emotional expression, and control strategies during the first year of life and infant attachment
outcomes at 12 months. Participants were middle-class Puerto Rican and Anglo mother–infant pairs
(N ϭ 60). Ratings of physical control, emotional expression, and maternal sensitivity during mother–
infant interactions in five everyday home settings, videotaped when the infants were 4, 8, and 12 months
old, were examined in combination with 12-month Strange Situation classifications. Results suggest that
physical control shows a different pattern of relatedness to maternal sensitivity, emotional expression,
and attachment outcomes among the Puerto Rican compared to the Anglo mothers in this study. These
findings have implications for practitioners and researchers interested in normative parenting among
diverse cultural groups.
RESUMEN: Esta investigacio´n se enfoca en las diferencias culturales en la relacio´n entre la sensibilidad
materna, la expresio´n emocional y las estrategias de control durante el primer an˜o de vida y los resultados
de la afectividad del infante a los 12 meses. En el estudio participaron 60 pares de madres-infantes de origen
puertorriquen˜o y anglosajo´n. Los puntajes de control fı´sico, expresio´n emocional y sensibilidad materna,
durante las interacciones entre la madre y su infante en cinco situaciones caseras de cada dı´a, grabadas
en vı´deo cuando los infantes tenı´an 4, 8 y 12 meses de nacidos, fueron examinados en combinacio´n con
las clasificaciones de la “Situacio´n Extran˜a” a los doce meses. Los resultados indican que el control fı´sico
muestra un patro´n diferente de relacio´n en cuanto a la sensibilidad materna, a la expresio´n emocional, y
This article is based on a doctoral dissertation, completed by the first author at the University of Connecticut in
December 2000. Portions of this article were presented at the 1999 and 2001 Biennial Meetings of the Society for
Research in Child Development. This research was made possible through a grant to the second author from the
National Institutes of Health (HD32800). We are grateful to Glorisa Canino of the University of Puerto Rico School
of Medicine for making available the resources of her laboratory for the collection of the Puerto Rican data. We would
also like to express appreciation to Margaret Adams, Eugenio Ayala, Delia Collazo, Jorge Colon, Olguimar Cruz, Xin
Feng, Zenaida Gonza´lez, Carmen Irizarry, Helena Mendez, Sylvia Meredith, Amy Miller, Erin Reutenauer, Pamela
Schulze, and Stephanie Wilson for their assistance with data collection and coding. Finally, we would like to thank
Preston A. Britner and Stephen A. Anderson for their invaluable comments on previous versions of this manuscript.
Direct correspondence to: Vivian J. Carlson, Department of Child Study, Education and Special Education, Saint
Joseph College, 1678 Asylum Avenue, West Hartford, CT 06117; phone: 860-231-5341; fax: 860-231-1691; e-mail:
[email protected].
INFANT MENTAL HEALTH JOURNAL, Vol. 24(1), 53– 73 (2003)
ᮊ 2003 Michigan Association for Infant Mental Health
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/imhj.10043
53
short
standard
base of drop
IMHJ (Wiley)
54
●
LEFT
BATCH
top of rh
base of rh
V.J. Carlson and R.L. Harwood
a los resultados de afectividad entre las madres de origen puertorriquen˜o, al compara´rseles con las madres
de origen anglosajo´n en este estudio. Estos resultados tienen implicaciones para los practicantes clı´nicos
y los investigadores interesados en la crianza normativa entre grupos culturales diversos.
cap height
base of text
RE´SUME´: L’accent de cette enqueˆte est place´ sur les diffe´rences culturelles dans la relation entre la
sensibilite´ maternelle, l’expression e´motionnelle et les strate´gies de controˆle durant la premie`re anne´e de
la vie et les re´sultats en matie`re d’attachement a` l’aˆge de 12 mois. Les participants consistaient en paires
me`re-nourrisson portoricaines et ame´ricaines, issues de la classe moyenne (N ϭ 60). Les e´valuations du
controˆle physique, de l’expression e´motionnelle et de la sensibilite´ maternelle durant les interactions
me`re-nourrison dans cinq situations quotidiennes a` la maison, filme´es a` la vide´o quand les be´be´s avaient
4, 8, et 12 mois ont e´te´ examine´es en association avec les classifications de situation e´trange pour l’aˆge
de 12 mois. Les re´sultats sugge`rent que le controˆle physique montre un pattern diffe´rent de relation a` la
sensibilite´ maternelle, a` l’expression e´motionnelle et aux re´sultats d’attachement chez les me`res Portoricains, compare´es dans cette e´tudes a` des me`res ame´ricaines. Ces re´sultats ont des implications pour les
praticiens et les chercheurs inte´resse´s par le parentage normatif au sein de groupes culturels divers.
ZUSAMMENFASSUNG: Diese Untersuchung untersucht kulturelle Differenzen in der Beziehung zwischen
mu¨tterlicher Feinfu¨hligkeit, Gefu¨hlsausdruck und Kontrolle wa¨hrend des ersten Lebensjahrs und die Bindung des Kleinkinds mit 12 Monaten. Die Teilnehmer waren puertorikanische und angloamerikanische
Mutter-Kind Paare aus der Mittelschicht (Nϭ60). Es wurden die Fa¨higkeit zur Ko¨rperkontrolle, der
Gefu¨hlsausdruck und die mu¨tterliche Feinfu¨hligkeit wa¨hrend der Mutter - Kind Interaktion in fu¨nf Alltagssituationen zu Hause mit Video beobachtet. Dies geschah im Alter des Kleinkinds von vier, acht und
zwo¨lf Monaten und wurde im Alter von zwo¨lf Monaten mit der Fremden Situation kombiniert. Die
Ergebnisse weisen darauf hin, daß das Muster der Korrelation zwischen der Fa¨higkeit des Kleinkinds
seinen Ko¨rper zu beherrschen, der mu¨tterlichen Feinfu¨hligkeit, dem Gefu¨hlsausdruck und der Beziehung
bei den puertorikanischen im Vergleich und den angloamerikanischen Mu¨ttern anders ist. Diese Ergebnisse haben Bedeutung fu¨r Praktiker und Forscher, die sich fu¨r normale Elternschaft bei verschiedenen
Kulturen interessieren.
*
*
*
INTRODUCTION
Attachment and Maternal Sensitivity
Attachment theory predicts that sensitive, responsive maternal care will lead to the development
of secure attachment relationships and subsequent socio-emotional competence (Ainsworth,
Blehar, Waters, & Wall, 1978; Bowlby, 1969). John Bowlby summarized the ideally sensitive
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
55
mother as “. . . being readily available, sensitive to her child’s signals, and lovingly responsive
when he or she seeks protection and/or comfort and/or assistance” (Bowlby, 1988, p. 4). This
description of sensitive caregiving closely parallels Ainsworth’s four caregiving dimensions of
sensitivity vs. insensitivity, acceptance vs. rejection, cooperation vs. interference, and accessibility versus ignoring (Ainsworth, Bell, & Stayton, 1974). These dimensions are based upon
both her naturalistic studies and her work using the Strange Situation to investigate patterns of
individual difference in attachment (Ainsworth, 1967; Ainsworth & Bell, 1970).
The hypothesis that sensitive, responsive maternal care is the single most important precursor of secure attachment relationships and subsequent socio-emotional competence (Ainsworth et al., 1974; Ainsworth et al., 1978; Bowlby, 1969) is now supported by meta-analyses
that find weak to moderate relationships between maternal sensitivity and security of attachment. However, the most recent meta-analysis (De Wolff & van IJzendoorn, 1997) found that
global assessments of emotional expression among nonclinical samples showed moderate effect
sizes equal to those of sensitivity ratings in predicting attachment quality. Such meta-analytic
findings are more compatible with a hypothesis in which maternal sensitivity is one of several
important precursors of secure attachment instead of the single, primary precursor.
Indeed, DeWolff and van IJzendoorn conclude that efforts to understand the specific contributions of sensitivity should be abandoned in favor of a multidimensional approach to investigating the antecedents of attachment. In particular, the role of parental management and
control strategies as well as broader contextual influences are highlighted as areas in need of
further investigation in studies of attachment antecedents (De Wolff & van IJzendoorn, 1997).
In addition, despite Ainsworth’s careful efforts to operationalize the concept of sensitivity,
several researchers have suggested that observer judgments regarding the “appropriateness” of
maternal responses are based in part on cultural and other contextual factors, and have recommended adding measures of a variety of maternal and contextual variables (level of social
support, personal attachment history/integration, cultural beliefs about parenting and child development) to investigations of the caregiving antecedents of attachment quality (Barnett, Kidwell, & Leung, 1998; De Wolff & van IJzendoorn, 1997; Fisher, Jackson, & Villarruel, 1998;
Harwood, Miller, & Irizarry, 1995; Jackson, 1986; Leyendecker, Lamb, Scho¨lmerich, & Fracasso, 1995; Rothbaum, Weisz, Pott, Miyake, & Morelli, 2000; Waters, Bailey, Pederson, &
Moran, 1999). Indeed, Ainsworth’s meticulous naturalistic studies in Uganda led her to caution
future researchers that caregiving must be understood in cultural and historic context (Ainsworth, 1967).
The most comprehensive effort to reframe the definition of sensitivity is presented by
George and Solomon (1999). Bowlby (1969) proposed that the infant’s developing attachment
behavioral system is supported by the attachment figure’s organized caregiving behavioral
system. Bowlby’s work, however, remained focused on the infant attachment system and little
or no investigation of the caregiving system was undertaken. George and Solomon suggest that
the effort to define sensitivity in terms of discrete behaviors or behavioral patterns be abandoned
in favor of an ethologically based behavioral systems approach to caregiving from the perspective of the attachment figure, not the child. The goal of the caregiving behavioral system
is the provision of protection from events or situations that the parent perceives as dangerous
or stressful. This system, like the child’s attachment system, is associated with strong emotions.
Mothers are intensely satisfied and pleased when they are successful at protecting their children
and similarly intensely negatively aroused when protective efforts are unsuccessful. The system
is activated or terminated by infant cues, internal processes, and/or environmental events. Optimal caregiving as an outcome of a behavioral system would by definition be organized by
caregiver cognitive processes, involve interactions with other behavioral systems both within
the caregiver and between the caregiver and others, and incorporate complex transactions be-
top of rh
base of rh
cap height
base of text
short
standard
IMHJ (Wiley)
56
●
LEFT
BATCH
V.J. Carlson and R.L. Harwood
tween biologic and experiential factors. The organization of a maternal caregiving system that
promotes the formation of secure attachments would thus include infant participation in the
feedback loops that activate and terminate the system, the caregiver’s mental representations
of attachment experiences, the influences of the family system on the mother – child relationship, and the broader context of cultural beliefs and values. Whereas Ainsworth’s definition of
sensitivity views caregiving behavior only from the child’s perspective, George and Solomon
challenge us to examine optimal caregiving from the perspective of the caregiver using the
same behavioral control systems approach which forms the foundation of attachment theory.
top of rh
base of rh
cap height
base of text
Physical Control and Maternal Sensitivity
Attachment research has traditionally focused on the dimensions of warmth and sensitive responsiveness with little attention given to issues of discipline and control (De Wolff & van
IJzendoorn, 1997). The Ainsworth et al. (1978) maternal sensitivity scales include the dimension of “cooperation versus interference.” The operational definitions of sensitivity thus include
reference to offering acceptable alternatives to infant requests that cannot be accommodated.
Efforts to control the infant’s behavior or to shape interactions based upon the mother’s wishes
are rated as interfering and insensitive. Thus maternal sensitivity has come to include the value
of “following the child’s lead,” or allowing the child to guide maternal behavior and make a
variety of choices. This aspect of sensitivity has been highlighted and oversimplified in the
context of the middle-class Anglo-American cultural emphasis on individual autonomy. [The
term “Anglo” has been chosen to describe the white American women of non-Hispanic European ancestry who participated in this study. Alternative labels such as “Euro-American” or
“U.S.” ignore the fact that many Puerto Ricans are of (Spanish) European ancestry and are also
all U.S. citizens. The cultural term “Anglo” has a long history of use to distinguish the Englishspeaking Americas from the Spanish-speaking Americas, and is far more parsimonious than
“white American women of non-Hispanic European ancestry” (Harwood, Scho¨lmerich, Ventura-Cook, Schulze, & Wilson, 1996).] Without regard to Ainsworth’s (1967) caution about
understanding the cultural and historic contexts of caregiving, the belief that control is always
intrusive and associated with negative outcomes appears to have been accepted as a universal
truth in Anglo-American parental ethnotheories.
This conceptualization of maternal control as interfering and insensitive is supported by
some studies using the Strange Situation which link high levels of maternal control with low
emotional availability or rejection to explain the child’s avoidant behavior when the attachment
system is activated. These studies find that high levels of maternal control in interactions are
experienced as overstimulating and intrusive and are associated with avoidant Strange Situation
behavior at 12 months (Ainsworth et al., 1978). In an investigation of the interface of attachment
theory and differential emotions theory, Malatesta, Culver, Tesman, and Shepard (1989) found
that infants classified as insecure-avoidant demonstrate expressions of heightened vigilance and
suppressed anger during Strange Situation reunions. These emotions are interpreted as infant
responses to aversive and overstimulating maternal interactions.
However, in studies among a variety of cultural groups which emphasize interdependence
over individual autonomy, there is evidence that the concept of sensitive maternal care includes
the expectation that mothers will structure and guide the infant’s environment and behavior to
enhance appropriate social behaviors and family relationships (Harwood, Scho¨lmerich,
Schulze, & Gonza´lez, 1999; Jones & Rao, 1998; Knight, Virdin, & Roosa, 1994; Leyendecker
& Lamb, 1999; Yang & Chang, 1999). In a cultural group that strongly values respectful,
appropriate behavior in the eyes of the community, the sensitive mother will consistently direct
her child’s attention to the needs of others and carefully guide or control his/her behavior in a
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
57
variety of social contexts (Harwood et al., 1995). Persistent physical control and strong limitations placed on infant behavior might be seen in this context not as interfering with the
infant’s development of autonomy, but as positive evidence of efforts to raise a well-behaved,
respectful child (Harwood, 1992). This type of maternal control, when combined with warm
and responsive parent – child relationships, appears to be associated with positive developmental outcomes among cultural groups characterized as more sociocentric than mainstream American culture (Barnett et al., 1998; Grusec, Hastings, & Mammone, 1994; Smith & Krohn, 1995).
It would thus appear that maternal control in the context of cultural meaning systems is an
important element of maternal sensitivity.
top of rh
base of rh
cap height
base of text
Parenting and Control
The relations between parental management and control strategies and child socio-emotional
outcomes are also examined in parenting research. These studies primarily utilize Baumrind’s
(1989, 1991) model of authoritarian, authoritative, permissive, and rejecting/neglecting parenting. Baumrind’s description of the optimal balance of responsiveness and demandingness
closely parallels Ainsworth’s (1974) definition of appropriate responsiveness as neither overstimulating nor understimulating in terms of the infant’s needs. This conceptual overlap between sensitivity and parenting style is the basis for much similarity among studies of the
antecedents and child outcomes of parenting styles and the antecedents and child outcomes of
sensitive caregiving. Indeed, parenting researchers have called for similar systemic approaches
to the investigation of cultural and individual beliefs and attributions in relation to parenting
practices as are found in the current attachment literature (Brody & Flor, 1998; Darling &
Steinberg, 1993; Deater-Deckard & Dodge, 1997; Greif & D’Agostino, 1999; Jones & Rao,
1998; Smetana, 1994; Spieker, Larson, Lewis, Keller, & Gilchrist, 1999).
In particular, Smetana (1994) has advocated adding the domains of warmth, coerciveness,
firm control, and communicative quality to Baumrind’s responsiveness-demandingness typology. The addition of these domains gives a broader context to descriptions of parenting styles
and leads to a wider variety of possible combinations of characteristics. One such combination
of parenting domains receiving attention in both the attachment and parenting literature is the
warm, communicative, but highly controlling parenting found among a variety of cultures
characterized as more interdependent than mainstream American culture. This style of parenting
combines the responsive, positive relationships typical of Baumrind’s authoritative style with
the high levels of control found in the authoritarian style. Jones and Rao (1998) use the term
“training” parenting to describe this style among Chinese parents. Brody and Flor (1998) use
“no nonsense” parenting to describe a similar style observed among African American parents.
Deater-Deckard and Dodge (1997) also offer evidence that the effects of physical discipline
on children are variable based upon the cultural context of the parent – child relationship. In
particular, these authors find that authoritarian discipline, when combined with a warm parentchild relationship, does not predict behavior problems among their African American participants. A recent report by Spieker et al. (1999) supports these findings among African Americans
and extends them to include negative verbal control strategies as well as harsh physical discipline. Rohner, Bourque, and Elordi (1996) also find that physical discipline is related to
maladjustment among White and African American low-income children only in the context
of a rejecting parental relationship.
Organizing these findings into an integrative model in which the emotional quality of the
parent – child relationship moderates parental control practices and child developmental outcomes has been addressed in the parenting literature (Darling & Steinberg, 1993). These authors
propose that understanding the ways in which parenting style moderates practices and child
short
standard
IMHJ (Wiley)
58
●
LEFT
BATCH
V.J. Carlson and R.L. Harwood
outcomes requires careful investigation of socialization goals, parenting practices, and the overall quality of the parent – child relationship. In this model, parenting style is defined as the
overall, expressed emotional climate of the family, and parenting practices are seen as specific,
goal-directed behaviors. Use of such an integrated, contextual model will assist in future explorations of the ways in which parenting styles and child outcomes vary among different
cultural groups, and how the influence of parenting styles and practices vary across the life
span.
top of rh
base of rh
cap height
base of text
Culture
Cultural meanings influence the interpretation of events and experiences, provide direction for
behavior, and evoke particular feelings (D’Andrade, 1984). The cultural meaning systems of
caregivers could thus be expected to include mental representations of ideal parent and child
behaviors and desirable child developmental endpoints, as well as to direct daily caregiving
behaviors, and evoke strong emotions in relation to childrearing beliefs and practices. Current
cultural research has also led to an increased understanding of culture as multidimensional as
opposed to a single contextual variable (Garcı´a Coll & Magnusson, 1999; Harwood, Handwerker, Scho¨lmerich, & Leyendecker, 2000).
Investigations of the development of attachment have only recently begun to attend to the
caregiving system from a control systems perspective, including cultural variations in parental
control strategies. Of these, very few investigations have focused on culture and parenting
among Latinos, who are projected to become 25% of American citizens by the year 2050
(Census Bureau, 1993). Understanding normative developmental pathways, childrearing beliefs, and parenting practices among Latinos is of definitive importance to this growing segment
of our future citizens.
In addition, many studies designed to investigate Puerto Rican and other Latino cultures
have focused on participants whose experiences are confounded by issues of immigration,
poverty, low educational levels, and unsafe living conditions (Fisher et al., 1998; Garcı´a Coll,
Meyer, & Brillon, 1995; Gonza´lez-Ramos, Zayas, & Cohen, 1998; Harwood, Leyendecker,
Carlson, Asencio, & Miller, 2002). Few investigations exist of normative, middle-class Puerto
Rican cultural beliefs and parenting practices.
Previous research (Harwood, 1992; Harwood et al., 1995) among lower and middle-class
Anglo and Puerto Rican mothers has suggested that mothers from these cultural groups perceive
and evaluate infant behaviors differently based on differential cultural emphases on self-maximization versus proper demeanor. In addition, these cultural beliefs have been related to differences among Anglo and Puerto Rican mothers in the structuring of mother – infant interactions in a variety of everyday contexts. In particular, Harwood et al. (1999) offer evidence that
middle-class island Puerto Rican mothers spend more time actively structuring the behavior of
their 12 – 15-month-old children than do middle-class Anglo mothers. This interactional style
is consistent with a greater cultural emphasis on proper demeanor and interdependence among
Puerto Rican mothers, but may appear to be controlling or intrusive when viewed from a
middle-class Anglo perspective.
Contributions of the Present Investigation
This study is expected to clarify some of the pathways by which the cultural patterning of early
experiences contribute to the development of maternal control strategies and mother – infant
attachment among a normative, middle-class group of Anglo and Puerto Rican mother – infant
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
59
pairs. In particular, the Puerto Rican mothers are expected to evidence higher levels of physical
control in the context of daily mother – infant interactions during the first year of life compared
to the Anglo mothers in this low-risk, middle-class sample. In addition, it is anticipated that
high ratings of physical control will be associated with different patterns of sensitivity, emotional expression, 12-month attachment classifications, and maternal reports of toddler socioemotional competence between the two cultural groups.
Evidence in support of these expected findings will provide additional documentation for
the necessity of further investigating the interface of culture and development. Research has
found evidence that cultural meanings are differentially assigned to attachment behaviors and
caregiving responses among a variety of diverse cultures (Friedlmeier & Trommsdorff, 1998;
Jackson, 1986; Morelli, Tronick, & Beeghly, 1999; Posada et al., 1995; Volker, Yovsi, &
Keller, 1998; Yang & Chang, 1999). In particular, cultural groups that differ along the continuum of sociocentric versus individualistic values demonstrate significant differences in parenting behaviors, patterns of attachment classifications, and child outcomes. Such differing
cultural meanings are likely to lead to differences in perceptions regarding what constitutes
sensitive, responsive caregiving, and optimal child outcomes.
Practically, this study seeks to identify alternate pathways to developmental competence
among Puerto Ricans compared to the familiar pathways based upon mainstream Anglo-American cultural values. Knowledge and understanding of desirable developmental end points,
normative parental expectations and practices, and their association with child socio-emotional
outcomes among these middle-class Puerto Ricans will enrich our ability to negotiate successful
interactions in schools and communities as well as inform understandings of our own cultural
limitations. Such knowledge will assist human service professionals in their efforts to provide
culturally competent programming that is respectful of cultural values, affirming of ethnic
identities, and supportive of inclusive participation in modern American life.
top of rh
base of rh
cap height
base of text
METHOD
In light of the fact that understanding the cultural context requires a complex investigation of
both culture-specific (emic) and theoretically universal (etic) concepts, this study utilizes a
“derived etic” approach (Berry, 1989). In other words, multiple methods and measures are
employed in an iterative effort to illuminate emerging commonalities and differences among
study participants (see also Harwood, Scho¨lmerich, & Schulze, 2000).
Sample
This study is part of a larger investigation of childrearing beliefs and practices across the first
year of life among middle-class Anglo-American mothers living in eastern Connecticut, and
middle-class Puerto Rican mothers living in San Juan, Puerto Rico. Middle-class Connecticut
Anglo (n ϭ 32) and island Puerto Rican (n ϭ 28) mothers and their infants were included in
the investigation. Because both socio-economic status and culture contribute to differences in
mothers’ childrearing beliefs and practices, the population includes only middle-class mothers.
Socio-economic status was defined using the Hollingshead (1975) four-factor scale, which
weights maternal and paternal education and occupational prestige into scores ranging from 8
(e.g., unskilled laborer, elementary school education only) to 66 (e.g., major professional with
graduate training). In this study, “middle-class” status was defined as maternal education beyond high school, with a household Hollingshead score above 40 (i.e., within Levels I and II
of the Hollingshead scale). Anglo mothers (n ϭ 32) were born, reared, and educated in the
short
standard
IMHJ (Wiley)
60
●
LEFT
BATCH
top of rh
base of rh
V.J. Carlson and R.L. Harwood
mainland United States, and spoke American English as their first language; Puerto Rican
mothers (n ϭ 27) were born and reared primarily in Puerto Rico, and spoke Puerto Rican
Spanish as their first language.
The study participants were recruited as volunteers through their local private pediatricians’
offices. Each mother had a four-month-old infant when enrolled in the study and agreed to
participate in home visits at 4, 8, and 12 months and a laboratory-based Strange Situation visit
at 12 months based on standard informed consent procedures as required by the University of
Connecticut and approved by the National Institutes of Health. Gender distribution was roughly
equal in both groups, and mothers were similar on a variety of demographic characteristics
(see Table 1).
cap height
base of text
Puerto Rican mothers. Among the Puerto Rican mothers, 46.4% identified themselves as
bilingual Spanish-English. Seven of the mothers had spent some time in the mainland U.S.
(mean ϭ 2.9 years, range one to seven years). To gain an index of psychologic acculturation
to the United States, Puerto Rican mothers completed a standardized instrument developed
specifically for Puerto Ricans by Garcı´a Coll and her colleagues (Tropp, Erkut, Alarco´n, Garcı´a
Coll, & Va´zquez, 1994). The instrument contains 31 items measuring cultural competence,
behaviors, preferences, and identification. The possible range for the total scale is a low of 31
(high Puerto Rican identity) to a high of 279 (high Anglo identity). The Puerto Rican mothers
TABLE 1. Mean Values for Sociodemographic Variables
Anglo
Puerto Rican
Mother’s age (years)
Father’s age (years)
Mother’s education (years)
Father’s education (years)
32.2
33.6
15.8
16.1
30.1
31.9
15.9
16.0
Percent mothers employed outside home
4 Months
8 Months
12 Months
67.6
62.5
62.5
78.6
78.6
85.7
*
Hours worked per week among employed mothers
4 Months
8 Months
12 Months
22.5
27.3
26.5
36.2
35.5
36.1
**
*
**
3.5
3.8
Total number children in household
Percent firstborn
Percent male
1.5
53.1
46.9
1.6
46.4
53.6
Percent mothers married
Percent nuclear family
90.6
87.5
82.1
82.1
Household Hollingshead score
51.4
51.3
Religious background (percent)
None
Protestant
Jewish
Catholic
31.3
25.0
6.3
37.5
7.1
32.1
0.0
60.7
Total number people in household
** p Ͻ .01.
p
*
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
61
in this study fell into the lower half of this scale (mean ϭ 84.4, range 41 – 142), indicating
high Puerto Rican identity.
top of rh
base of rh
cap height
base of text
Procedure
All mothers were interviewed in their homes by ethnically matched interviewers when their
target infant was 4, 8, and 12 months of age. Mothers were interviewed in their native language:
Puerto Rican Spanish or American English. At each time point, mothers were interviewed
regarding a variety of parental beliefs and practices, and videotaped for about 45 minutes
interacting with their infants in five everyday situations (feeding, bathing/dressing, social play,
object-oriented teaching, and free play). A standardized laboratory-based Strange Situation
procedure was performed within a week following the 12-month home visit.
Procedures to promote crosscultural validity. Translations of the English protocols into Spanish were provided by bilingual, bicultural Puerto Rican graduate research assistants at the
University of Connecticut and reviewed by the principal investigator of the original study in
consultation with two Puerto Rican cultural consultants. All protocols were examined for cultural suitability by the primary investigator of the original study and the Puerto Rican cultural
consultants. The Spanish translations were then checked for preservation of meaning and cultural appropriateness by the bicultural, bilingual Puerto Rican graduate assistants and the original primary investigator. All protocols were pilot tested with members of each cultural group
to ensure that they were readily comprehensible and ethnographically valid.
Ratings of maternal behavior. The 4-, 8-, and 12-month videotapes of mother – infant interactions were used to rate maternal behavior on three scales: maternal sensitivity, emotional
expression, and physical control. The maternal behavior rating scales utilized Ainsworth’s
(1978) nine-point maternal sensitivity scale. In addition, emotional expression and physical
control scales were developed by the first author after an extensive literature review and in
consultation with the principal investigator and a graduate student colleague. Because of their
predictive power in previous meta-analytic studies (DeWolff & van IJzendoorn, 1997; Goldsmith & Alansky, 1987), maternal sensitivity and emotional expression were selected as variables to assess the traditional relationships between sensitive caregiving and Strange Situation
classifications in this normative sample. Physical control was included as a means of assessing
the potentially different cultural relations among control, sensitivity, positive affect, and attachment outcomes. The emotional expression and physical control scales use the same ninepoint Likert scale format as Ainsworth’s maternal sensitivity scales, with a rating of one as the
negative or lowest expression of the trait and a rating of nine as the most positive or highest
rating possible. Every effort was made to operationalize the author-developed scales using
observable physical and vocal behaviors and to limit the need for observer judgements regarding “appropriateness” based on contextual cues. The emotional expression scale includes observations of facial expression, voice quality, and body language (Ainsworth et al., 1978;
Biringen & Robinson, 1991; Field, 1994; Jones, Davalos, & Lundy, 1999) in ratings that range
from “negative” (one) to “delighted” (nine). Physical control observations rate the mother’s
observed use of physical contact to manipulate, limit, or control the infant’s movements on a
scale ranging from “almost no physical control” (one) to “nearly constant physical control”
(nine).
Strange Situation. The Strange Situation (Ainsworth & Wittig, 1969) is a laboratory-based
procedure developed to assess security of attachment. The setting is a waiting room area with
short
standard
IMHJ (Wiley)
62
●
LEFT
BATCH
top of rh
base of rh
V.J. Carlson and R.L. Harwood
adult chairs and infant toys. The room is videotaped through a one-way mirror as the mother
and a stranger enter the room, leave the room, and interact with the infant. Attachment security
classifications are based upon the infant’s reunion behaviors when the mother returns after two
brief separations. The second author, who attended a week-long Strange Situation workshop
at the University of Minnesota taught by Alan Sroufe, trained graduate students in Connecticut
and Puerto Rico with regard to the proper administration of the Strange Situation.
cap height
base of text
Reliability. The maternal behavior rating scales were coded by ethnically matched graduate
assistants trained to criterion by the first author. Ratings of maternal behavior during individual
activity segments (feeding, bathing/dressing, social play, free play and teaching) at each time
point (4, 8, and 12 months) were recorded on forms developed for this study. These graduate
assistants were blind to the study hypotheses and achieved a final interrater reliability correlation of r ϭ .91 (p Ͻ .01) (range ϭ .86 to .94) using equal numbers of randomly chosen
Anglo and Puerto Rican tapes to establish initial reliability and to recheck reliability during
the coding at the 4-, 8-, and 12-month time points for a total of 20% of the sample. The Strange
Situation videotapes were coded by external coders at the University of Washington, trained
to criterion and blind to the study hypotheses.
Telephone follow-up survey. A follow-up study of a subgroup of these mothers explored
potential group differences in toddler socio-emotional outcomes of 12-month Strange Situation
classifications. A standardized telephone interview using a modified version of the InfantToddler Social and Emotional Assessment (ITSEA) maternal report scale (Carter & BriggsGowan, 1997) was conducted with 27 of the Anglo and 15 of the Puerto Rican mothers. The
ITSEA was shortened in consultation with its authors to include those child behaviors most
salient to attachment status.
RESULTS
Strange Situation Classifications
Strange Situation classification results indicated similar percentages of securely attached infants
among both cultural groups (59% Anglo, 52% Puerto Rican). Insecurely attached infants,
however, showed unexpected percentages in both groups. Thirty percent of the Puerto Rican
infants were classified as insecure avoidant and 22% of Anglo infants were classified as insecure
ambivalent (see Table 2). Attachment research among middle-class Anglo infants usually reports an avoidant classification rate of about 20% and an ambivalent classification rate of about
10%.
TABLE 2. Percent Strange Situation Classifications at 12 Months
Classification
Anglo (n ϭ 32)
Puerto Rican (n ϭ 27)
Avoidant (A)
Secure (B)
Ambivalent (C)
Disorganized (D)
12.5% (4)
59.4% (19)
21.9% (7)
6.3% (2)
29.6% (8)
51.9% (14)
7.4% (2)
11.1% (3)
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
63
Socio-Demographic Variables and 12-Month
Attachment Classification
top of rh
base of rh
cap height
base of text
To explore the possible contribution of socio-demographic factors to these unexpected patterns
of 12-month attachment classifications, analyses of variance for the continuous scale variables
and chi-square analyses for the nominal variables were performed across both cultural groups
and for each group. Nine scale demographic variables (maternal and paternal age, education,
and Hollingshead scores; household Hollingshead; total number of children; and total number
of people in the household) were entered together into ANOVAs to explore potential differences between avoidant/not avoidant, ambivalent/not ambivalent, and secure/insecure attachment status. These comparisons yielded no significant relations. Chi-square analyses of the six
nominal demographic variables (child gender, birth order, marital status, religious affiliation,
child care, and geographic stability) by avoidant/not avoidant, ambivalent/not ambivalent, and
secure/insecure attachment status also yielded no significant relations.
Ratings of Maternal Physical Control
To reduce the large number of independent variables (three rating scales ϫ five settings ϫ
three time points) in relation to the sample size (Anglo n ϭ 32; Puerto Rican n ϭ 27), the
observational settings were grouped into goal-oriented activities that require high levels of
mother – child cooperation (i.e., feeding and teaching) and open-ended situations (i.e., free play,
social play, and bathing/dressing). Bathing/dressing was included in the open-ended group
because the middle-class mothers in this study all approached the bathing/dressing task in a
playful, unstructured manner requiring minimal cooperation from the infant. Mean ratings for
these structured and open-ended situations were computed for each mother at each time point.
The correlation matrix for these grouped rating scales indicates that no correlations met the .90
criteria for variable exclusion based on multicollinearity issues (Tabachnick & Fidell, 1996,
p. 84).
To test the hypothesis that, Puerto Rican mothers would evidence higher overall rates of
physical control, but similar overall ratings of maternal sensitivity and emotional expression
when compared to Anglo mothers, means were computed for each mother’s ratings of sensitivity, emotional expression, and physical control across both situations and all time points.
Doubly multivariate repeated measures analyses of variance were conducted for open and feed/
teach emotional expression, maternal sensitivity, and physical control at each time point by
cultural group. When maternal sensitivity, emotional expression, and physical control are examined together, results show significant differences between cultural groups, F(6, 52) ϭ 3.51,
p Ͻ .01, and a significant time by group interaction, F(12, 46) ϭ 2.13, p Ͻ .05. There were
no main effects of time. Tests of between subjects effects indicate that only open-ended and
feed/teach physical control differed significantly by cultural group, F(1, 57) ϭ 7.86, p Ͻ .01;
F(1,57) ϭ 21.4, p Ͻ .001, respectively. The time by group interaction was significant only for
open-ended physical control, F(2, 114) ϭ 6.27, p Ͻ .01, and open-ended emotional expression,
F(2, 114) ϭ 3.43, p Ͻ .05.
Follow-up one-way ANOVA’s examined the influence of cultural group membership on
ratings of open-ended and feed/teach physical control at each of the three time points. Puerto
Rican mothers were rated as significantly higher in the use of feed/teach physical control at 4
months, F(1, 57) ϭ 8.13, p Ͻ .01, at 8 months, F(1, 57) ϭ 10.27, p Ͻ .01, and at 12 months,
F(1, 57) ϭ 17.27, p Ͻ .001, compared to Anglo mothers. Puerto Rican mothers were rated as
significantly higher in the use of open-ended physical control at 8 months, F(1, 57) ϭ 29.43,
p Ͻ .001, and at 12 months, F(1, 57) ϭ 4.19, p Ͻ .05, compared to Anglo mothers. A follow-
short
standard
IMHJ (Wiley)
64
●
LEFT
BATCH
V.J. Carlson and R.L. Harwood
up ANOVA showed no significant differences by cultural group for open-ended emotional
expression at any of the time points.
top of rh
base of rh
cap height
base of text
Maternal Physical Control, Sensitivity, Emotional
Expression, and Infant Attachment Classifications
Similar rates of secure attachment combined with evidence of significant differences in maternal
physical control, but not maternal sensitivity or emotional expression between these groups,
suggests potentially differential patterning of these attributes among the Puerto Rican and
Anglo mother – infant pairs. In particular, the traditional links between high control, low sensitivity, and insecure avoidant attachment may not be applicable to the Puerto Rican mothers
and their infants.
Tests of the second hypothesis, that associations between maternal physical control, maternal sensitivity, emotional expression, and 12-month infant attachment classifications would
be different for each cultural group, were conducted using discriminant function analyses
(DFA). DFA is appropriate for use with naturally occurring groups of unequal size (i.e., attachment classification groups) and allows in depth exploration of the nature of group differences as opposed to identifying statistically significant differences (as in MANOVA).
MANOVAs were used to select variables for inclusion in the discriminant function analyses for each separate cultural group. This process included all 18 of the open and feed/teach
ratings on the three maternal behavior scales at 4, 8, and 12 months (three rating scales ϫ two
settings ϫ three time points). These data were entered as independent variables to examine
their potential association with the 12-month attachment classification groups (avoidant, secure,
ambivalent, and disorganized). The analyses were conducted separately for each cultural group.
Variables showing significance, or a trend for significance in relation to the attachment classifications were noted for each group: 12-month open physical control, F(3, 23) ϭ 2.71, p Ͻ
.10, 4-month open emotional expression, F(3, 23) ϭ 2.57, p Ͻ .10, and 12-month feed/teach
maternal sensitivity, F(3, 23) ϭ 3.01, p Ͻ .05 for the Puerto Rican group; and 8-month feed/
teach emotional expression, F(3, 28) ϭ 2.30, p Ͻ .10 and 12-month open physical control,
F(3, 28) ϭ 3.36, p Ͻ .05 for the Anglo group. Twelve-month feed/teach physical control was
also selected for the Anglo group based upon knowledge of its shared variance with the previously selected 12-month open physical control variable, the theoretical salience of contiguous
variables, and the current study emphasis on the role of physical control.
DFAs were then computed for the Anglo group using the three selected variables (physical
control in open-ended and feed/teach tasks at 12 months; and emotional expression in feed/
teach tasks at 8 months) to discriminate three attachment classification groups. The two infants
classified as disorganized for the Anglo group resulted in one cell of this DF with fewer cases
than independent variables and were eliminated from the analysis. This analysis produced two
functions, significant at the p Ͻ .05 level. The first function accounted for 52.4% of the variance. Box’s M test was not significant at the .05 alpha level, providing evidence that the
assumption of homogeneity of variance had not been violated in this analysis. Classification
results for the 30 Anglo (two cases eliminated as discussed above) dyads indicate that 77%
were classified correctly when prior probabilities for functions were calculated based on actual
group size for each attachment classification (see Table 3).
The patterns of prediction evident in these discriminant functions for the Anglo variables
show the expected consistent association between high levels of control, high levels of emotional expression in goal-oriented tasks, and avoidant attachment (Table 4).
The DFA for the Puerto Rican group using the three selected rating scale variables (openended emotional expression at 4 months; open-ended physical control at 12 months; and feed/
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
65
top of rh
base of rh
cap height
base of text
TABLE 3. Strange Situation Discriminant Function Analysis Classification Rates for the
Anglo Group (n ϭ 30)
Predicted Classifications
Actual Child Classification
Secure (B)
Avoidant (A)
Ambivalent (C)
Secure (B)
Avoidant (A)
Ambivalent (C)
18
2
3
0
1
0
1
1
4
Correct classifications: 77%.
teach maternal sensitivity at 12 months) produced three DF’s significant at the p Ͻ .01 level,
with the first function accounting for 59% of the variance. When the Box’s M alpha level is
set at .01 to correct for small sample size and unequal groups, the test is not significant,
providing evidence for the overall validity of the analysis (Tabachnick & Fidell, 1996). Classification results for these 27 Puerto Rican dyads indicated that 74% were classified correctly
when prior probabilities for functions were calculated based on actual group size for each
attachment classification (see Table 5).
Patterns of prediction evident in the Puerto Rican discriminant function variables showed
an association between high levels of 12-month physical control and secure attachment; high
levels of four-month emotional expression and avoidant attachment; and high levels of 12month sensitivity and ambivalent attachment (see Table 6).
A follow-up study of a subgroup of these mothers explored potential group differences in
toddler socio-emotional outcomes of 12-month Strange Situation classifications. A standardized
telephone interview using a maternal report scale (Infant-Toddler Social and Emotional Assessment (ITSEA); Carter & Briggs-Gowan, 1997) was conducted with 27 of the Anglo and
15 of the Puerto Rican mothers. Puerto Rican mothers reported significantly more internalizing
behaviors across all attachment classifications as compared to the Anglo mothers (respective
means ϭ 8.00, 5.07, p Ͻ .05). High levels of reported internalizing behaviors were associated
with ambivalent attachment among the Anglo, but not the Puerto Rican toddlers. In addition,
regression models of this data find that the Depression/Social Withdrawal and Aggression/
Defiance maternal report subscales significantly predict avoidant/not avoidant status for the
Anglo group, F(3, 23) ϭ 4.92, p Ͻ .01; however, no subscales significantly predict avoidant/
not avoidant status for the Puerto Rican group. A study examining the concurrent validity of
the ITSEA measure among mainland U.S. middle-class mother – infant pairs found patterns of
TABLE 4. DFA Anglo (n ϭ 30) Patterns of Attachment Prediction: Rating Variable Means
High
Moderate
Low
Open-ended physical control
12 Months
(A) avoidant
(B) secure
(C) ambivalent
Feed/teach physical control
12 Months
(A) avoidant
(B) secure
(C) ambivalent
Feed/teach emotional expression
8 Months
(A) avoidant
(C) ambivalent
(B) secure
short
standard
IMHJ (Wiley)
66
●
LEFT
BATCH
top of rh
base of rh
V.J. Carlson and R.L. Harwood
TABLE 5. Strange Situation Discriminant Function Analysis Classification Rates for the Puerto Rican
Group (n ϭ 27)
cap height
base of text
Predicted Classification
Actual Child
Classification
Secure (B)
Avoidant (A)
Ambivalent (C)
Disorganized (D)
Secure (B)
Avoidant (A)
Ambivalent (C)
Disorganized (D)
12
4
0
1
1
4
0
0
1
0
2
0
0
0
0
2
Correct classifications: 74%
association between ITSEA scales and attachment quality similar to these Anglo results (Carter,
Little, Briggs-Gowan, & Kogan, 1999).
DISCUSSION
These results call into question the use of a single, universal definition of maternal sensitivity,
instead providing evidence that sensitive caregiving behaviors may be culturally constructed,
incorporating the socialization goals, values, and beliefs of the family and community. In
particular, the expected link between high levels of maternal control and insecure attachment
is not found among these Puerto Rican participants. The Puerto Rican mothers, compared to
the Anglo mothers, used significantly more physical control overall in their interactions with
their infants during the first year of life. These mothers persistently and actively structured
interactions with their infants in a manner consistent with their long-term socialization goals
(Harwood et al., 1999). Typically, attachment theory has conceptualized this type of control
as interfering and has rated such behavior as insensitive (Ainsworth et al., 1974). However,
our findings suggest that the highest ratings of physical control were associated with secure
12-month attachment status for these middle-class Puerto Rican dyads. This apparently paradoxical finding highlights the need for culturally specific definitions of sensitive caregiving. In
contrast, attachment theory predictions based upon the traditional (middle-class Anglo) definition of sensitive caregiving are supported by the results among our Anglo mother – infant
TABLE 6. DFA Puerto Rican (n ϭ 27) Patterns of Attachment Prediction: Rating Variable Means
Open-ended physical control
12 Months
Open-ended emotional expression
4 Months
Feed/teach maternal sensitivity
12 Months
High
Moderate
Low
(B) secure
(D)
(A) Avoidant
(C) ambivalent
(A) avoidant
(D)
(C) ambivalent
(B) secure
(C) ambivalent
(B) secure
(D)
(A) avoidant
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
67
pairs — that is, high use of maternal physical control was associated with insecure avoidant
attachment.
Previous research demonstrates a coherent relationship between childrearing beliefs and
behaviors in everyday contexts for these middle-class Anglo and Puerto Rican mothers (Harwood, et al., 1999). Teaching infants to be attentive, calm, and well-behaved requires considerably more physical prompting and control than teaching infants to be assertive and selfconfident. Thus, it appears that maternal use of physical control may be regulated by maternal
socialization goals in a meaningful, predictable manner among these Puerto Rican and Anglo
participants.
In response to the DeWolf and van IJzendoorn (1997) meta-analytic findings that ratings
of emotional expression among low-risk, middle class samples are nearly equal in effect size
to ratings of maternal sensitivity, this study uses global ratings of emotional expression to
measure warmth in maternal – infant interactions. Dividing these everyday interactions into
goal-directed tasks that require considerable infant cooperation for successful completion versus more open-ended, playful interactions enables clearer distinctions between “appropriate”
levels of maternal emotional expression. The highest ratings of emotional expression in structured tasks at eight months are associated with avoidant attachment among the Anglo dyads.
The infant may experience this high level of emotional expression as intrusive and distracting
in the context of a routine, structured task, leading to an association with avoidant strategies
as predicted by previous research (Belsky, Rovine, & Taylor, 1984; Isabella & Belsky, 1991;
Malatesta et al., 1989; Pederson & Moran, 1996). High use of emotional expression by the
Puerto Rican mothers in open-ended tasks at 4-months is also associated with avoidant attachment at 12-months. The very limited self-regulation capacities of a four month old would
indicate the need for a calm, soothing maternal demeanor in playful interactions. Intense emotional expression directed toward such a young infant may be experienced as disruptive and
may thus tend to promote infant withdrawal and preference for self-directed regulation strategies. Some support may be found for this interpretation in the findings reported by Kogan
and Carter (1996) showing an association between infant negativity in the four-month stillface procedure and maternal intrusiveness. Malatesta et al. (1989) also hypothesize that high
levels of maternal emotional contingency in close interactions are experienced as aversive by
the infant and tend to promote disengagement. In addition, Cohn, Campbell, and Ross (1991)
find that infant negativity during the still-face procedure predicts avoidant attachment at 12
months.
No ratings of maternal sensitivity were included as significant predictors of attachment
status for the Anglo group. Variability in maternal sensitivity ratings is quite limited among
these middle-class, low risk, Anglo dyads. Indeed, both feed/teach and open maternal sensitivity
mean ratings show only small effect sizes in relation to secure/insecure attachment status
r 2 ϭ .03 and .04, respectively) for this small Anglo sample (Cohen, 1988). The maternal
sensitivity and emotional expression ratings are moderately intercorrelated (r ϭ .37 to .49,
p Ͻ .01), but show no significant correlations with physical control ratings. For these participants, the greater variability in use of physical control and emotional expression (as discussed
above) appear to provide stronger predictions of attachment status than maternal sensitivity
ratings.
One 12-month maternal sensitivity rating was included as a significant predictor in the
discriminant function analysis for the Puerto Rican group. Interestingly, the highest ratings of
maternal sensitivity in structured tasks at 12 months are associated with ambivalent attachment
among the Puerto Rican dyads. A similar finding is reported by Scho¨lmerich, Lamb, Leyendecker, and Fracasso (1997) in an examination of mother – infant teaching interactions and
top of rh
base of rh
cap height
base of text
short
standard
IMHJ (Wiley)
68
●
LEFT
BATCH
V.J. Carlson and R.L. Harwood
attachment security. Ambivalent dyads were found by Scho¨lmerich et al. to have the highest
levels of coordination of social attention in teaching tasks, particularly in the person-oriented
category. Although the authors do not speculate about possible reasons for this finding, it may
be that these infants are already exhibiting early forms of high social vigilance in response to
inconsistent maternal responsiveness. The lowest ratings of maternal sensitivity among the
Puerto Rican mothers are associated with avoidant attachment, consistent with attachment
theory predictions.
An additional puzzle exists in the unexpected pattern of attachment classifications across
the two cultural groups in this study. These results, showing nearly 30% of Puerto Rican infants
classified as avoidant, are counterintuitive given the Puerto Rican emphasis on interdependence
and familial closeness. In a comparison of attachment classification across cultures, Fracasso,
Busch-Rossnagel, and Fisher (1994) report a similar 30% avoidant classification rate among
low-income Puerto Rican and Dominican participants living in New York City. These authors
speculate that such unexpected distributions of infant attachment classifications are related to
low socio-economic status, gender effects (with males showing a more typical pattern of attachment classifications), and cultural differences in parenting behaviors (security was associated with more sensitivity and more abrupt pickups). Avoidant classification has been traditionally interpreted as a psychologic defense for coping with a lack of maternal warmth
(Ainsworth et al., 1974). For participants in this study, overall mean ratings of maternal sensitivity and emotional expression across the first year of life do not differ significantly by
cultural group, and no gender effects are found, providing evidence that the higher incidence
of avoidant attachments among Puerto Rican infants is not significantly related to these factors.
One possible explanation is that the higher percentage of employed Puerto Rican mothers
at 12 months may have increased the incidence of avoidant attachment among the Puerto Rican
infants. Research by Clarke-Stewart (1989) and Belsky and Rovine (1988) has shown a weak
but significant association between maternal employment and avoidance in the Strange Situation. These studies include meta-analytic data obtained from diverse samples of more than
1200 participants, and thus may be relevant to the cultural contexts examined in the present
study. Compared to Anglo mothers, Puerto Rican mothers in this study were significantly more
likely to be employed when their infants were 12 months of age. In addition, employed Puerto
Rican mothers worked significantly more hours than Anglo mothers at all time points during
the first year of life (see Table 1). However, additional analyses show no significant differences
for attachment classification by maternal hours worked per week for either cultural group.
To further explore the effects of the family system as well as cultural beliefs and values
on the maternal caregiving system, associations among the full range of socio-demographic
factors and 12-month attachment classifications were explored. Again, no significant relationships were found. Thus, we find no evidence that the measured socio-demographic factors have
a significant impact on attachment relationships among these middle-class mother – infant pairs.
This finding most likely reflects the purposeful limitation of socio-demographic variability
inherent in the original study design.
It is also possible that some of the avoidant classifications among this group of low-risk,
middle-class Puerto Rican infants may represent an avoidance of maternal physical control in
preference for unfettered exploration of a new environment. This avoidant response would not
necessarily be associated with a rejecting or negative maternal emotional relationship. Some
support for this pattern may be found in a study by Lewis and Feiring (1989), in which evidence
is presented for an association between avoidance in the Strange Situation at 12 months and a
preference for exploration present in early infancy. The fact that there is a 50% incorrect secure
classifications rate among the Puerto Rican avoidant infants in this study (see Table 5) may
reflect just such an avoidance of maternal control among these four infants. Their mothers’
top of rh
base of rh
cap height
base of text
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
69
emotional expression and maternal sensitivity ratings would not be expected to differ from
those of mothers with securely attached infants, resulting in the incorrect discriminant function
classifications. If further evidence is found for this pattern of response, it may represent a
different developmental pathway in the context of normative socialization goals and practices
for Puerto Rican mothers and their infants. Such an alternative pathway would not be expected
to predict the type of negative outcomes sometimes associated with avoidant attachment among
Anglo infants.
Preliminary data from a 28-month follow-up study with a subgroup of these mother –
infant pairs finds evidence for cultural differences in caregiver perceptions of ideal child behaviors. Internalizing behaviors such as wanting to be on mother’s lap and crying or clinging
when mother leaves are consistent with the Puerto Rican emphasis on appropriate relatedness
and interdependence, but inconsistent with the Anglo emphasis on confidence and independence. These data also offer limited support for the finding that avoidant Strange Situation
behavior is associated with different developmental outcomes among the Anglo compared to
the Puerto Rican toddlers (Carlson & Harwood, 1999). Although this follow-up study is seriously limited by small group size (many families had moved since the 12-month contact and
could not be located), these findings are consistent with the hypothesis of a pathway whereby
differential cultural patterning of caregiving behaviors affects child socio-emotional development.
Expanding our definitions of maternal sensitivity to include a variety of culture-specific
beliefs and practices does not alter the underlying association between sensitive caregiving and
secure attachment. Ainsworth et al. (1974) conclude that the most fundamental characteristic
of sensitivity is the mother’s ability to establish a harmonious relationship with her infant. It
may be that harmonious maternal – infant relationships are best established by caregiving practices designed to produce culturally valued traits in the growing infant. In the case of these
Puerto Rican mothers, high levels of control in the context of warm, affectionate relationships
may best serve to encourage obedience, respect, and relatedness in their children.
Conceptualizing caregiving as a behavioral system enables understanding of the need for
both universality and specificity in our definitions of optimal caregiving. The universal goal of
providing protection from perceived threats forms the foundation of optimal caregiving. The
nature of perceived threats is universal only in regard to immediate threats to physical wellbeing and survival. All other less tangible threats are culturally constructed and may vary widely
from one culture to the next. For example, child behavior patterns that are inconsistent with
culturally valued behaviors are threats to the child’s developing social competence and acceptance in the larger community. Thus, clingy, dependent behavior may elicit protective responses and efforts to control or change these behaviors on the part of Anglo parents who
value individual autonomy. On the other hand, clingy, dependent behavior may be accepted
by the Puerto Rican mothers as evidence of appropriate familial interdependence.
The caregiving system framework also highlights the need to include a variety of parent,
child, and contextual influences in investigations of the development of attachment security. If
the caregiving system is investigated using the same behavioral systems approach as attachment
theory, then the role of culture must be seen as central to the caregiver’s mental representations
and interpretations of relationships and experiences. Although the goal of providing protection
from situations the parent views as stressful or dangerous remains universal, perceptions of
danger or stress and means of protection will vary widely as a function of cultural meaning
systems, contextual, and experiential differences. Evidence of the ways in which caregiver
socialization goals and parenting practices differentially influence child developmental outcomes among these Puerto Rican and Anglo mother – infant pairs provides preliminary support
for the cultural specificity needed in definitions of optimal caregiving.
top of rh
base of rh
cap height
base of text
short
standard
IMHJ (Wiley)
70
●
LEFT
BATCH
top of rh
base of rh
V.J. Carlson and R.L. Harwood
Further explorations of the role of culture in early relationship formation, caregiving patterns, and child development will require longitudinal evaluations of child outcomes for each
cultural group. In addition, investigations into the ways in which migration and acculturation
affect caregiving values and behaviors across generations should also provide more in-depth
understanding of the interface between culture and development. Culturally varied precursors
of attachment as discussed above require further validation, as do any potential cultural variations in the socio-emotional outcomes of attachment security among different groups. Attention to the caregiving behavioral system from the point of view of the caregiver will enable
investigation of both universal caregiving goals and culturally specific definitions of optimal
parenting and ideal child outcomes.
cap height
base of text
REFERENCES
Ainsworth, M.D.S. (1967). Infancy in Uganda: Infant care and the growth of love. Baltimore, MD: Johns
Hopkins University Press.
Ainsworth, M.D.S., & Bell, S.M. (1970). Attachment, exploration, and separation: Individual differences
in strange-situation behavior of one-year-olds. Child Development, 41, 49–67.
Ainsworth, M.D.S., Bell, S.M., & Stayton, D.J. (1974). Infant–mother attachment and social development: Socialisation as a product of reciprocal responsiveness to signals. In M.J.M. Richards (Ed.),
The integration of a child into a social world (pp. 99–135). London: Cambridge University Press.
Ainsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns of attachment: A psychological
study of the strange situation. Hillsdale, NJ: Erlbaum.
Ainwsorth, M., & Wittig, B. (1969). Attachment and exploratory behavior of one-year-olds in a strange
situation. In B. Foss (Ed.), Determinants of infant behavior (vol. 4). New York: Barnes & Noble.
Barnett, D., Kidwell, S.L., & Leung, K.H. (1998). Parenting and preschooler attachment among lowincome urban African American families. Child Development, 69, 1657–1671.
Baumrind, D. (1989). Rearing competent children. In W. Damon (Ed.), Child development today and
tomorrow. San Francisco: Jossey-Bass.
Baumrind, D. (1991). Parenting style and adolescent development. In J. Brooks-Gunn & A.C. Petersen
(Eds.), The encyclopedia of adolescence (pp. 746–758). New York: Garland.
Belsky, J., & Rovine, M.J. (1988). Nonmaternal care in the first year of life and the security of infant–
parent attachment. Child Development, 59, 157–167.
Belsky, J., Rovine, M., & Taylor, D.G. (1984). The Pennsylvania infant and family development project,
III. The origins of individual differences in infant-mother attachment: Maternal and infant contributions. Child Development, 55, 718– 728.
Berry, J.W. (1989). Imposed etics-emics-derived etics: The operationalization of a compelling idea. International Journal of Psychology, 24, 721–735.
Biringen, Z., & Robinson, J. (1991). Emotional availability in mother–child interactions: A reconceptualization for research. American Journal of Orthopsychiatry, 61, 258–271.
Bowlby, J. (1969). Attachment. New York: Basic Books.
Bowlby, J. (1988). Developmental psychiatry comes of age. The American Journal of Psychiatry, 145,
1–10.
Brody, G., & Flor, D.L. (1998). Maternal resources, parenting practices, and child competence in rural,
single-parent African American families. Child Development, 69, 803–816.
Carlson, V.J., & Harwood, R.L. (1999). Culture and strange situation behavior at 12 months: Anglo and
Puerto Rican mothers’ reports of socioemotional development. Paper presented at the Biennial Meeting of the Society for Research in Child Development, Albuquerque, NM.
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
71
Carter, A.S., & Briggs-Gowan, M. (1997). The Infant–Toddler Social and Emotional Assessment (ITSEA). Unpublished measure. New Haven, CT: Yale University.
top of rh
base of rh
cap height
base of text
Carter, A.S., Little, C., Briggs-Gowan, M.J., & Kogan, N. (1999). The infant–toddler social and emotional
assessment (ITSEA): Comparing parent ratings to laboratory observations of task mastery, emotion
regulation, coping behaviors, and attachment status. Infant Mental Health Journal, 20, 375–392.
Clarke-Stewart, K.A. (1989). Infant day care. Maligned or malignant? American Psychologist, 44, 266–
273.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence
Erlbaum.
Cohn, J., Campbell, S., & Ross, S. (1991). Infant response in the still-face paradigm at 6 months predicts
avoidant and secure attachment at 12 months. Development and Psychopathology, 3, 367–376.
D’Andrade, R.G. (1984). Cultural meaning systems. In R. Shweder & R. LeVine (Eds.), Culture theory:
Essays on mind, self, and emotion (pp. 88–117). New York: Cambridge University Press.
Darling, N., & Steinberg, L. (1993). Parenting style as context: An integrative model. Psychological
Bulletin, 113, 487– 496.
Deater-Deckard, K., & Dodge, K.A. (1997). Externalizing behavior problems and discipline revisited:
Nonlinear effects and variation by culture, context, and gender. Psychological Inquiry, 8, 161–175.
DeWolff, M.S., & van IJzendoorn, M.H. (1997). Sensitivity and attachment: A meta-analysis on parental
antecedents of infant attachment. Child Development, 68, 571–591.
Field, T. (1994). The effects of mother’s physical and emotional unavailability on emotion regulation.
Monographs of the Society for Research in Child Development, 59, 250–283.
Fisher, C.B., Jackson, J.F., & Villarruel, F.A. (1998). The study of African American and Latin American
children and youth. In R. Lerner (Ed.), The handbook of child psychology (5th ed., Vol. 1, pp. 1145–
1207). New York: John Wiley and Sons.
Fracasso, M.P., Busch-Rossnagel, N.A., & Fisher, C.B. (1994). The relationship of maternal behavior
and acculturation to the quality of attachment in Hispanic infants living in New York City. Hispanic
Journal of Behavioral Sciences, 16, 143–154.
Friedlmeier, W., & Trommsdorff, G. (1998). Mother–child interaction in Japan and Germany: Its relation
to development of emotion regulation. Paper presented at the XVth Biennial Meetings of the International Society for the Study of Behavioral Development, Bern, Switzerland.
Garcı´a Coll, C., & Magnusson, K. (1999). Cultural influences on child development: Are we ready for a
paradigm shift? In A.S. Masten (Ed.), Cultural processes in child development (Vol. 29, pp. 1–24).
Mahwah, NJ: Erlbaum.
Garcı´a Coll, C., Meyer, E.C., & Brillon, L. (1995). Ethnic and minority parenting. In M.H. Bornstein
(Ed.), Handbook of parenting: Vol. 2. Biology and ecology of parenting (pp. 189–209). Mahwah,
NJ: Erlbaum.
George, C., & Solomon, J. (1999). Attachment and caregiving. The caregiving behavioral system. In J.
Cassidy & P.R. Shaver (Eds.), Handbook of attachment. Theory, research, and clinical applications
(pp. 649– 687). New York: The Guilford Press.
Goldsmith, H.H., & Alansky, J.A. (1987). Maternal and infant predictors of attachment: A meta-analytic
review. Journal of Consulting and Clinical Psychology, 55, 805–816.
Gonza´lez-Ramos, G., Zayas, L.H., & Cohen, E.V. (1998). Child-rearing values of low-income, urban
Puerto Rican mothers of preschool children. Professional Psychology: Research and Practice, 29,
377– 382.
Greif, M.L., & D’Agostino, N.M. (1999). The impact of maternal warmth, protection, and control on
children’s socioemotional functioning. Paper presented at the Biennial Meeting of the Society for
Research in Child Development, Albuquerque, NM.
Grusec, J.E., Hastings, P., & Mammone, N. (1994). Parenting cognitions and relationship schemas. In
J.G. Smetana (Ed.), Beliefs about parenting: Origins and developmental implications (Vol. 66, pp.
5–18). San Francisco: Jossey-Bass.
short
standard
IMHJ (Wiley)
72
●
LEFT
BATCH
V.J. Carlson and R.L. Harwood
Harwood, R.L. (1992). The influence of culturally derived values on Anglo and Puerto Rican mothers’
perceptions of attachment behavior. Child Development, 63, 822–839.
top of rh
base of rh
cap height
base of text
Harwood, R.L., Handwerker, W.P., Scho¨lmerich, A., & Leyendecker, B. (2001). Ethnic category labels
and the contextualized individual: An exploration of the individualism/sociocentrism debate. Parenting: Science and Practice, 1, 217– 236.
Harwood, R.L., Leyendecker, B., Carlson, V., Ascencio, M., & Miller, A.M. (2002). Parenting among
Latino families in the U.S. In M. Bornstein (Ed.), The handbook of parenting: Vol. 4. Social conditions and applied parenting (2nd ed.). Mahwah, NJ: Erlbaum.
Harwood, R.L., Miller, J.G., & Irizarry, N.L. (1995). Culture and attachment: Perceptions of the child in
context. New York: Guilford Press.
Harwood, R.L., Scho¨lmerich, A., & Schulze, P.A. (2000). Homogeneity and heterogeneity in cultural
belief systems. In W. Damon (Series Ed.) & S. Harkness, C. Raeff, & C.M. Super (vol. Eds.), New
directions for child and adolescent development: Variability in the social construction of the child
(87) (pp. 41– 57). San Francisco: Jossey-Bass.
Harwood, R.L., Scho¨lmerich, A., Schulze, P.A., & Gonza´lez, Z. (1999). Cultural differences in maternal
beliefs and behaviors: A study of middle-class Anglo and Puerto Rican mother–infant pairs in four
everyday situations. Child Development, 70, 1005–1016.
Harwood, R.L., Scho¨lmerich, A., Ventura-Cook, E., Schulze, P.A., & Wilson, S.P. (1996). Culture and
class influences on Anglo and Puerto Rican mothers’ beliefs regarding long-term socialization goals
and child behavior. Child Development, 67, 2446–2461.
Hollingshead, A.B. (1975). Four factor index of social status. Unpublished manuscript, Yale University.
Isabella, R.A., & Belsky, J. (1991). Interactional synchrony and the origins of infant–mother attachment:
A replication study. Child Development, 62, 373–384.
Jackson, J.F. (1986). Characteristics of black infant attachment behaviors. The American Journal of Social
Psychiatry, VI, 32– 35.
Jones, E., & Rao, M. (1998). The differential impact of parental beliefs and practices on Chinese and
English preschoolers’ social competence. Paper presented at the XV Biennial Meetings of the International Society for the Study of Behavioural Development, Bern, Switzerland.
Jones, N.A., Davalos, M., & Lundy, B. (1999). Depressed mothers’ ratings of mother/infant interactions.
Paper presented at the Biennial Meeting of the Society for Research in Child Development, Albuquerque, NM.
Knight, G.P., Virdin, L.M., & Roosa, M. (1994). Socialization and family correlates of mental health
outcomes among Hispanic and Anglo American children: Consideration of cross-ethnic scalar equivalence. Child Development, 65, 212– 224.
Kogan, N., & Carter, A. (1996). Mother– infant reengagement following the still face: The role of maternal
emotional availability in infant affect regulation. Infant Behavior and Development, 19, 359–370.
Lewis, M., & Feiring, C. (1989). Infant, mother, and mother–infant interaction behavior and subsequent
attachment. Child Development, 60, 831– 837.
Leyendecker, B., & Lamb, M.E. (1999). Latino families. In M.E. Lamb (Ed.), Parenting and child development in “nontraditional” families (pp. 247–262). Mahwah, NJ: Erlbaum.
Leyendecker, B., Lamb, M.E., Scho¨lmerich, A., & Fracasso, M.P. (1995). The social worlds of 8- and
12-month-old infants: Early experiences in two subcultural contexts. Social Development, 4, 194–
208.
Malatesta, C.Z., Culver, C., Tesman, J.R., & Shepard, B. (1989). The development of emotion expression
during the first two years of life. Monographs of the Society for Research in Child Development,
54(1–2), 55– 82.
Morelli, G.A., Tronick, E., & Beeghly, M. (1999). Is there security in numbers? Child care in a hunting
and gathering community and infants’ attachment relationships. Paper presented at the Biennial
Meetings of the Society for Research in Child Development, Albuquerque, NM.
short
standard
IMHJ (Wiley)
RIGHT
BATCH
Attachment, Culture, and the Caregiving System
●
73
Pederson, D.R., & Moran, G. (1996). Expressions of the attachment relationship outside of the strange
situation. Child Development, 67, 915–927.
top of rh
base of rh
cap height
base of text
Posada, G., Gao, Y., Wu, F., Posada, R., Tascon, M., Scho¨lmerich, A., Sagi, A., Kondo-Ikemura, K.,
Haaland, W., & Synnevaag, B. (1995). The secure-base phenomenon across cultures: Children’s
behavior, mothers’ preferences, and experts’ concepts. Monographs of the Society for Research in
Child Development, 60, 27– 48.
Rohner, R.P., Bourque, S.L., & Elordi, C.A. (1996). Children’s perceptions of corporal punishment,
caretaker acceptance, and psychological adjustment in a poor, biracial Southern community. Journal
of Marriage and the Family, 58, 842– 852.
Rothbaum, F., Weisz, J., Pott, M., Miyake, K., & Morelli, G. (2000). Attachment and culture: Security
in the United States and Japan. American Psychologist, 55, 1093–1104.
Scho¨lmerich, A., Lamb, M.E., Leyendecker, B., & Fracasso, M.P. (1997). Mother–infant teaching interactions and attachment security in Euro-American and Central-American immigrant families. Infant
Behavior and Development, 20, 165– 174.
Smetana, J.G. (1994). Parenting styles and beliefs about parental authority. New Directions for Child
Development, 66, 21– 36.
Smith, C., & Krohn, M.D. (1995). Delinquency and family life among male adolescents: The role of
ethnicity. Journal of Youth and Adolescence, 24, 69–93.
Spieker, S.J., Larson, N.C., Lewis, S.M., Keller, T.E., & Gilchrist, L. (1999). Developmental trajectories
of disruptive behavior problems in preschool children of adolescent mothers. Child Development,
70, 443– 458.
Tabachnick, B.G., & Fidell, L.S. (Eds.). (1996). Using multivariate statistics (3rd ed.). New York:
HarperCollins.
Tropp, L.R., Erkut, S., Alarco´n, O., Garcı´a Coll, C., & Va´zquez, H. (1994). Toward a theoretical model
of psychological acculturation. Center for Research on Women, Working Papers Series No. 268.
Wellesley College: Wellesley, MA.
U.S. Census Bureau. (1993). Hispanic Americans today (Vol. P23-183). Washington, DC: Government
Printing Office.
Volker, S., Yovsi, R., & Keller, H. (1998). Maternal interactional quality as assessed by non-trained raters
from different cultural backgrounds. Paper presented at the XVth Biennial ISSBD Meetings, Bern,
Switzerland.
Waters, C.A., Bailey, H.N., Pederson, D.R., & Moran, G. (1999). A “person-focused” approach to identifying and describing patterns of maternal behavior. Paper presented at the Biennial Meetings of
the Society for Research in Child Development, Albuqurque, NM.
Yang, Y.S., & Chang, M.J. (1999). The relationships of mothers’ parenting and children’s characteristics
to Korean preschool children’s attachment and social competence. Paper presented at the Society
for Research in Child Development Biennial Meeting, Albuquerque, NM.
short
standard