Bell and Richard’s position is very similar to Mac-Donald’s (1992) argument that attachment involves atleast two independent systems. The first is a fear-basedsystem that is essentially the same as that originally de-scribed by Bowlby (1969), and the second is awarmth-based system that is very similar to thecaregiving system described by Bell and Richard. Bothsystems of attachment, as related to the parent–childrelationship, can be understood in terms of parental in-vestment (Trivers, 1972, 1974), as is described in thefirst section. The second section addresses a few mis-conceptions about evolution and proximate and ulti-mate causes.
Parental Investment
Parent–child attachment, whether involving prox-imity-related systems or caregiving-warmth, is readilyunderstandable in terms of the more general topic ofparental investment; that is, parent–child attachment isa manifestation of the evolution of parental invest-ment. Trivers (1974) defined parental investment as re-sources (e.g., time, food, etc.) that are provided tooffspring at a cost to the parent; costs can be reproduc-tive (e.g., a delay in conceiving the next offspring) orin terms of the health and survival of the parent(Clutton-Brock, 1991). This is not to say that each andevery parental behavior is directly related to the par-ent’s survival or later reproduction. Rather, the sys-tems that orient parents to their offspring, as inmaintaining proximity and promoting parental invest-ment (e.g., in terms of time and material resources), arecostly to parents and would not evolve unless they pro-vided a reproductive benefit to parents. In nonhumanspecies, the primary reproductive benefit of parentalinvestment is an increase in offspring survival rates(Clutton-Brock, 1991).
An evolved bias to invest in offspring does not inany way mean that all parents at all times will invest intheir offspring (Daly & Wilson, 1981). For many non-human species, parents will often abandon offspring orotherwise reduce levels of parental investment, contin-
gent on current conditions; for instance, if food is tooscarce to ensure the survival or normal development ofoffspring, then parents will often abandon these off-spring (Clutton-Brock, 1991). Stated somewhat differ-ently, from an evolutionary perspective, parentalinvestment is expected to be expressed contingent onsocial (e.g., population density) and ecological (e.g.,availability of food) factors and on parental condition(e.g., health, social status). When parents invest in off-spring, some aspects of this investment are obligate,that is, essential for offspring survival, and other as-pects may be expressed more conditionally, orfacultatively. For the latter, the investment may not beessential for survival, but if provided, it results in otherbenefits to offspring, such as an improvement in latersocial, and thereby reproductive, competitiveness.
The long developmental period of humans and thenecessarily heavy cost of human parental investmentstrongly argue that this investment is an evolved fea-ture of human social behavior, as indicated by Bell andRichard and others (e.g., Bowlby, 1969; Lovejoy,1981; MacDonald, 1992). As with other species, it islikely that the selection pressure for human parental in-vestment was reduced child mortality risks and per-haps improvements in the later social, and therebyreproductive, competitiveness of children (Geary,2000). The proximity-seeking feature of attachmentdescribed by Bowlby (1969) may reflect a form of ob-ligate parental investment, whereas thecaregiving-warmth system described by Bell andRichard and by MacDonald (1992) may reflect a moreconditional form of parental investment.
Features of proximity seeking, such as offspringdistress calls, are found in many species of primate andmammal (Hofer, 1987) and reflect a way of solicitingparental investment that has a long evolutionary his-tory. For instance, many features of proximity seeking,such as stranger anxiety and separation anxiety, arefound in all human societies and are manifested in thesame way and appear at about the same age in one ofour closest relatives, the chimpanzee (Pan troglodytes;Bard, 1995; Eibl-Eibesfeldt, 1989). The apparently ho-
mologous attachment patterns in humans and chim-panzees suggest that this form of human parentalinvestment has at least a 6 million–year evolutionaryhistory, the estimated age of the ancestor common tohumans and chimpanzees (e.g., Vrba, 1985). In fact,the ubiquity of proximity-related features of attach-ment across many species and the increased mortalityrisks associated with separation from the primary care-giver (e.g., Goodall, 1986) strongly suggest that thisaspect of attachment predated hominid evolution.
The caregiving-warmth feature of parental invest-ment, in contrast, is not a feature of attachment in mostother mammals or primates and, in fact, is absent inmany human societies (Geary, 1998; MacDonald,1992). Although the proximate mechanisms are notfully understood, it appears that high levels of conflictwithin the wider society or within the family result inreductions in parental empathy and responsiveness tochildren and in harsh child-rearing practices (e.g.,painful physical punishment). At the same time, mostof these parents show the proximity-related behaviors(e.g., responding to distress signals) described byBowlby (1969; MacDonald, 1992). Although theseharsh child-rearing practices are sometimes describedas abusive and pathological, the possibility that theseare adaptive and facultatively expressed responses toharsh social conditions needs to be seriously consid-ered. In fact, MacDonald argued that harsh child rear-ing “shuts down” the neurobiological and associatedemotional systems that support interpersonal warmthand thus results in a more guarded, aggressive, and ex-ploitative social style. Although maladaptive in somecontexts, this type of social style is functional in othersocial contexts—in theory, those contexts that promoteharsh parental child rearing.
It appears that caregiving, as described by Bell andRichard, and warmer parent–child relations are morecommonly found in high-resource, low-conflict socialecologies and in settings with low mortality rates and arelatively even distribution of resources among the lo-cal population (Belsky, Steinberg, & Draper, 1991;MacDonald, 1992; Wilson & Daly, 1997). In these set-tings, a cooperative social style—one characterized bywarm and reciprocal social relationships amongadults—may be more adaptive than the guarded andaggressive social style described above. In this view,cooperative social relationships among adults, com-bined with living in a low-risk and high-resource envi-ronment, result in the facultative expression of thefeatures of caregiving described by Bell and Richard.This form of caregiving, in turn, results in a bias to-ward more reciprocal and warmer social relationships(MacDonald, 1992).
In short, the proximity-seeking features of attach-ment described by Bowlby (1969) appear to be an obli-gate feature of parental investment—althoughabandonment of children would still be expected under
some conditions (Daly & Wilson, 1981)—with a longevolutionary history. The caregiving-warmth featureof attachment is less common across species and, thus,likely has a much shorter evolutionary history. More-over, the human caregiving and warmth feature doesnot appear to be an obligate form of parental invest-ment in humans but rather appears to be expressedfacultatively, that is, only under certain social and eco-logical conditions (MacDonald, 1992).
Proximate and Ultimate Causes
The ultimate selection pressures for any form of pa-rental investment—obligate or conditional—are im-provements in offspring survival rates and, in somespecies, including humans, an improvement in the so-cial and later reproductive competitiveness of off-spring (Clutton-Brock, 1991; Geary, 2000). Theassociated proximate causes constitute all of thosecognitive, social, and emotional systems that result inthe administration of parental investment, independ-ently of whether this investment is subjectively com-fortable or cost free to the parent. Bell and Richardargue that “no rational self-interested person wouldvoluntarily pay the emotional and monetary costs ofparenting” (this issue). This statement represents aconfusion of ultimate and proximate mecha-nisms—even subjectively taxing proximate costs, suchas anxiety, distress, and lost resources, will evolve, ifthey result in a reproductive advantage(Clutton-Brock, 1991). In other words, genetic self-in-terest—of which parental investment is but one exam-ple—and here-and-now personal self-interest (e.g.,personal comfort) are not the same.
In fact, selection could support heavy emotionaland other costs of parental investment if these costs re-sulted in behaviors that reduced the mortality risks ofchildren. As an example, parental grief associated withthe loss of a child and the associated examination of“how did this happen” could prompt behavioralchanges that reduced risks to later children. Moreover,there are other mechanisms, such as fear when thechild is threatened, empathy as described by Bell andRichard, and child characteristics (e.g., cute) that pro-mote parental investment (Trivers, 1974). On a per-sonal level, this investment often exacts emotional andmonetary costs on the parent, to the benefit of the child.But if this investment increases the chances that thesechildren will reach adulthood and have children oftheir own, then parental investment is, from a geneticperspective, a self-interested behavior.
Note
David C. Geary, Department of Psychology, 210McAlester Hall, University of Missouri at Columbia, Co-lumbia, MO 65211–2500. E-mail: [email protected]
85
COMMENTARIES
References
Bard, K. A. (1995). Parenting in primates. In M. H. Bornstein (Ed.),Handbook of parenting: Vol. 2. Biology and ecology ofparenting(pp. 27–58). Mahwah, NJ: Lawrence Erlbaum Asso-ciates, Inc.
Belsky, J., Steinberg, L., & Draper, P. (1991). Childhood experience,interpersonal development, and reproductive strategy: An evo-lutionary theory of socialization.Child Development, 62,647–670.
Bowlby, J. (1969).Attachment and loss: Vol. 1. Attachment. London:Hogarth.
Clutton-Brock, T. H. (1991).The evolution of parental care. Prince-ton, NJ: Princeton University Press.
Daly, M., & Wilson, M. (1981). Abuse and neglect of children in evo-lutionary perspective. In R. D. Alexander & D. W. Tinkle(Eds.),Natural selection and social behavior(pp. 405–416).New York: Chiron Press.
Eibl-Eibesfeldt, I. (1989).Human ethology. New York: Aldine deGruyter.
Geary, D. C. (1998).Male, female: The evolution of human sex differ-ences. Washington, DC: American Psychological Association.
Geary, D. C. (2000). Evolution and proximate expression of humanpaternal investment.Psychological Bulletin, 126,55–77.
Goodall, J. (1986).The chimpanzees of Gombe: Patterns of behavior.Cambridge, MA: Belknap Press.
Hofer, M. A. (1987). Early social relationships: A psychobiologist’sview. Child Development, 58,633–647.
Lovejoy, C. O. (1981). The origin of man.Science, 211,341–350.MacDonald, K. (1992). Warmth as a developmental construct:
An evolutionary analysis. Child Development, 63,753–773.
Trivers, R. L. (1972). Parental investment and sexual selection. In B.Campbell (Ed.),Sexual selection and the descent of man1871–1971(pp. 136–179). Chicago: Aldine.
Trivers, R. L. (1974). Parent–offspring conflict.American Zoologist,14,249–264.
Vrba, E. S. (1985). Ecological and adaptive changes associated withearly hominid evolution. In E. Delson (Ed.),Ancestors: Thehard evidence(pp. 63–71). New York: Alan R. Liss.
Wilson, M., & Daly, M. (1997). Life expectancy, economic in-equality, homicide, and reproductive timing in Chicagoneighbourhoods. British Medical Journal, 314,1271–1274.
The Complexity of the Caregiving System:A Perspective From Attachment Theory
Jude CassidyDepartment of Psychology
University of Maryland
In one of his earliest writings, Bowlby (1956)pointed out that further understanding of attachmentcould be gained from examination of the mother’s tieto her infant. Although Bowlby (1984) later wrotebriefly about “parenting behavior” from a biologicalperspective as “like attachment behavior … in somedegree preprogrammed” (p. 271), he devoted his atten-tion almost exclusively to examining the attachmentside of what he called the “attachment-caregiving so-cial bond” (Bowlby, 1969/1982, p. 377). It is to hiscredit that his initial focus on the sole behavioral sys-tem of attachment produced such a highly generativetheory that subsequent theorists examining thecaregiving system have had a solid theoretical and em-pirical base on which to build (see Bretherton,Biringin, Ridgeway, Maslin, & Sherman, 1989;Bretherton, Golby, & Cho, 1997; Cassidy, 1999, inpress; Heard & Lake, 1997; Main, 1990; Pianta,Marvin, Britner, & Borowitz, 1996; Slade, Belsky,Aber, & Phelps, 1999; Zeanah, Benoit, Hirschberg,Barton, & Regan, 1994; and most extensively, George& Solomon, 1996, 1999, and Solomon & George,1996).
In their target article, Bell and Richard claim that“attachment theory oversimplifies and furthertrivializes the process of caregiving” (this issue). In
this article, I present a contrasting perspective andattempt to demonstrate contemporary attachment the-ory’s focus on the multifaceted, rich complexities in-volved in the caregiving behavioral system. I reviewdiscussion by a number of attachment theorists who ar-gue that the caregiving system should be examined notonly as a separate system, but also in connection withother parental behavioral systems and in relation to thechild’s behavioral systems. These attachment theoristsalso consider cognition and emotion, a variety of con-textual factors, both past and present influences oncaregiving, a life-span perspective, and differentialmaternal and paternal caregiving—all aspects of a the-oretical perspective that is anything but simple.
The Caregiving Behavioral System
From an attachment theory perspective, caregivingis considered within the framework of the caregivingbehavioral system. Bowlby (1969/1982) borrowed theconcept of the behavioral system from the ethologiststo describe a species-specific system of behaviors thatleads to certain predictable outcomes, at least one ofwhich contributes to the individual’s reproductive fit-ness. The predictable outcome of activation of the par-ent’s caregiving system is child–parent proximity, and
86
COMMENTARIES
the biological function is protection of the child andenhanced parental reproductive fitness. The concept ofthe behavioral system involves inherent motivation.There is no need to view attachment as the by-productof any more fundamental processes or “drives.” Thus,Bell and Richard’s claim that “attachment accounts ofcaregiving lack a plausible motivation for caregiving, areason why the caregiver should choose to nurture thedependent” (this issue) dismisses Bowlby’s initialthinking about the caregiving system and more recentexpansions of this thinking that are solidly within con-temporary biological theorizing and empirical re-search. The caregiving system is thought to haveevolved because, during the time when humans andother primates were initially evolving, parents whoprovided protective care to their offspring were morelikely to have offspring survive and pass on the genesfor this caregiving behavior. This notion of inherentmotivation is also compatible with Piaget’s (1954) for-mulation of the inherent motivation of the child’s inter-est in exploration.
Activation and Termination of theCaregiving Behavioral System
Bell and Richard claim that, according to attach-ment theory, “the caregiving behavioral system is inac-tive most of the time. … [The] behavioral controlsystem [described] by most attachment theorists … iseither turned on or turned off” (this issue). Bell andRichard contrast the “on-off caregiving of attachmenttheories” with their own perspective, wherein “caringis conceptualized as enduring and variable” (this is-sue). Yet, in parallel to Bowlby’s (1969/1982) revisedview of the attachment system as never fully turnedoff, the caregiving system is viewed by attachment the-orists as continually active (enduring) and variable(i.e., the interaction of a variety of factors contributesto whether caregiving behavior occurs at all and the in-tensity of that behavior). Bell and Richard furtherclaim that, from an attachment theory perspective, thecaregiving system “only becomes active when the de-pendent’s distress vocalizations or excessive distancefrom the caregiver stimulates it” (this issue). In fact,within the framework of attachment theory, the mecha-nism is not so simple. There are thought to be multipleand complex factors involved in the relative activationand deactivation of the caregiving system; activationof caregiving is viewed as far more complexly deter-mined than simple responsiveness to child attachmentbehavior:
As is true for many behavioral systems, activation ofthe caregiving system results from both internal andexternal cues. Internal cues include presence of hor-mones, cultural beliefs, parental state (e.g., whetherthe parent is tired or sick), and activation of other pa-
rental behavioral systems (e.g., exploration,food-getting, fear). External cues include state of theenvironment (e.g., whether it is familiar, whether thereis danger, whether others are present and who thoseothers are), state of the infant (e.g., whether the infantis sick or tired), and behavior of the infant (e.g.,whether he or she is exhibiting attachment behavior).(Cassidy, 1999, p. 10)
Interplay Among Behavioral Systems
Further adding to the complex nature of thecaregiving behavioral system is the fact that it oper-ates in conjunction with a variety of a parent’s otherbiologically based behavior systems. An appreciationof the interplay among behavioral systems has alwaysbeen central to attachment theory, and two examplesof this interplay involving the caregiving system arenoted here. For instance, just as the child’s attachmentand fear systems are closely linked (such that an in-crease in fear contributes to increased attachment be-havior; Bowlby, 1973), so are the parent’s caregivingand fear systems. When a parent is frightened bythreat to the child, she increases caregiving behavior;conversely, when a parent is in a safe environmentand perceives no threat, she is likely to decreasecaregiving behavior. Similarly, the exploratory andcaregiving systems are likely to be intertwined, suchthat activation of a parent’s exploratory system mayreduce activation of the caregiving system: On SuperBowl Sunday, a father may be more lax in hiscaregiving behavior during the game than followingit. (See George & Solomon, 1999, for a full discus-sion of “competing behavioral systems.”)
Parent–Child Conflict
The caregiving system operates not only in concertwith other parental behavioral systems, but also in con-cert with the child’s behavioral systems. Bowlby(1969/1982) considered the child’s attachment systemand the parent’s caregiving system as “complemen-tary” behavioral systems. In most cases, parents andchildren work together to maintain a comfortable, safedegree of proximity. If the child moves away, the par-ent will retrieve him; if the parent moves away, thechild will follow or signal for the parent to return. Fol-lowing Bowlby’s thinking, it seems likely that whenthe caregiving system is relatively activated, thechild’s attachment system can be relatively deacti-vated; attachment behaviors are not needed becausethe parent has assumed responsibility for maintainingproximity. If the caregiving system is not relatively ac-tivated, then the child’s attachment system becomesactivated, should the context call for it. This “dynamicequilibrium” (Bowlby, 1969/1982, p. 236) contributesto understanding the notion of the mother’s providing a
87
COMMENTARIES
“secure base from which to explore.” The mother’smonitoring of child–mother proximity frees the infantfrom such necessary monitoring and permits thechild’s greater attention to exploration (Main, 1973),although there are certainly times when both partners’systems are concurrently activated and both take sig-nificant responsibility for maintaining proximity.
However, as Trivers (1974) noted, parents and chil-dren are at times in conflict because circumstances en-hance the reproductive fitness of each member of thedyad differently. For instance, when an infant’s attach-ment system is activated in the presence of the mother,the infant’s sole wish is for her to respond. Althoughsuch infant behavior is usually a powerful activatingstimulus for the mother’s caregiving system, themother may choose among several competing needsand may or may not provide care. The child’s concernis immediate and focused; the mother’s concerns maybe more diffuse and long-range (Main, 1981, 1990).The mother may have to leave the infant to work tosupport the family, or she may have several children towhose needs she must attend. Parental caregiving to aparticular child is only one of many ways a parent canincrease reproductive fitness. Others include havingadditional children; providing care to other children;maintaining ties to a mate, kin, or peers who will helpprotect offspring; and gathering resources. Thus, par-ents juggle a variety of complex, competing factorswhen faced with caregiving decisions. Main (1990)has proposed that from an evolutionary perspective,maternal unresponsiveness may be useful to themother if it maximizes the total number of survivingoffspring (see also Belsky, 1999; Simpson, 1999).
The Role of Cognition
A central notion of attachment theory is that early inlife, humans develop mental representations of them-selves, the people important in their lives, and theworkings of the world (Bowlby, 1969/1982, 1973;Bretherton & Munholland, 1999; Main, Kaplan, &Cassidy, 1985). More recent attention to attachmentacross the life span has led to the proposition that par-ents also develop representations of themselves as par-ents and of their children (Slade et al., 1999; Zeanah etal., 1994), as well as a “current state of mind with re-gard to attachment” (Main & Goldwyn, 1998). Thethinking is that parental mental representations influ-ence parenting behavior, and there is evidence that thisis the case (e.g., Slade et al., 1999). Understanding theprocess through which this occurs (selective attention,memory, interpretative bias, expectations) will be im-portant in future research. What factors contribute tochange in representational models and the relative in-fluence of early and later models will also be importantto explore.
The Role of Emotion
The emotions linked to caregiving may be as pow-erful as any experienced throughout life. As I noted ina previous article (Cassidy, 1999), the birth of a firstchild (which establishes the adult as a parent) is oftenaccompanied by feelings of great joy; threats to thechild are accompanied by anxiety; the death of a childbrings profound grief. This intertwining of thecaregiving system with intense emotions may resultfrom selective pressures of evolution: Enhanced repro-ductive fitness may result when, for instance, a par-ent’s anxiety about threats to a child prompts the parentto seek effective interventions.
Explaining Parental Responsiveness
According to Bell and Richard, “attachment logichas suggested why sensitivity and responsiveness maybe important for the child, but it has not been able toexplain why they are important to the parent” (this is-sue). Yet consideration of the biological function of thecaregiving system—protection of the young and en-hanced parental reproductive fitness—is central to anattachment theory perspective, and such considerationleads to a biological understanding of why, for pri-mates, responsiveness and sensitivity are “important tothe parent.” The case of parental soothing serves as oneexample. Why would a parent expend effort to soothe achild frightened of thunder when the parent correctlyperceives that the child is in no danger? There are anumber of possible reasons, all of which increase pa-rental reproductive fitness. I have proposed elsewherethat soothing behaviors serve, indirectly, to facilitatethe parent’s monitoring of potential or real dangers tothe child (Cassidy, 1999). Parental provision of contactusually comforts a distressed child. If the child contin-ues to be distressed for a substantial time followingcontact, there may be another threat of which the par-ent in unaware. Through continuing attempts to soothethe child, the parent gains information about threat tothe child. The parent may not realize, for instance, thatthe child has a painful splinter in her foot. Furthermore,there are many ways in which inconsolable crying (be-yond infancy) can signal serious health problems; aparent will not know whether crying is inconsolableunless the parent attempts to console. Another possiblereason for soothing is that it may contribute to reducingthe health risks associated with chronic, unrelievedstress (e.g., Kobassa, 1979). In addition, it is likely thatthe child learns, through modeling, about how tosoothe the distress of others. Given evidence thatadults prefer romantic partners who are sensitively re-sponsive (Pietromonaco & Carnelley, 1994; Zeifman& Hazan, 1997), having learned such behavior may in-crease the child’s later ability to attract desirable ro-
88
COMMENTARIES
mantic (reproductive) partners. Both reducing risk tothe child’s health and enhancing the child’s attractive-ness to desirable partners, in turn, enhance the parent’sown reproductive fitness. Another reason, which isperhaps the most basic reason for soothing, may be thatthe species has evolved such that a child’s distress isaversive to the parent (even at the level of auditorypitch and frequency), and the parent desires and acts toterminate the aversive stimulus.
The Active Parent: UnderstandingIndividual Differences
Bell and Richard refer to “the essentially passivecaregiving behavioral control system described in at-tachment theory” wherein the parent “responds me-chanically to the child’s crying” (this issue).Behavioral systems, however, have never been viewedas “passive” systems by ethologists or attachment the-orists. Furthermore, there has been considerable recenttheorizing that individual differences in parentalcaregiving can be viewed in part as (nonconscious) ac-tive strategic parental attempts to increase reproduc-tive fitness. In other words, it has been proposed thatparents (nonconsciously) consider a complex constel-lation of factors and arrive at a (nonconscious) beliefabout the type of caregiving likely to enhance their re-productive fitness. This parental belief, in turn, islikely to influence the parents’ active choices ofcaregiving behaviors. Thus, although the parametersof the caregiving system are thought to be biologicallybased and to have evolved through the process of natu-ral selection, the caregiving system is thought to haveevolved with the capacity to flexibly adapt to environ-mental variation in ways that enhance reproductive fit-ness. For further discussion of the ways attachmenttheorists view the caregiving system as characterizedby active parental judgment and learning rather than aseither “passive” or “strictly innate,” as Bell and Rich-ard claim, see Belsky (1999), George and Solomon(1999), and Main (1990).
The Influence of Past and Present
There has long been a theoretical view that past ex-periences guide present parenting. Fraiberg (1980), forinstance, claimed, “we do unto others as we were doneto.” Similarly, Sroufe and Fleeson (1986) pointed outthat babies learn both sides of important early dyadicrelationships—at the same time they learn what it islike to be a baby in a relationship, they learn how it isthat the caregiver is. Bowlby (1979) also proposed thatsecure childhood attachment would facilitate an indi-vidual’s later parenting. More recently, theorists havebegun to examine the complex intertwining of past and
present influences on parenting: What are the relativeinfluences of past and present? Under what circum-stances is one more influential? How much of a protec-tive factor can early positive experience provide? Dorelative influences vary across individuals? What is therole of biological change during adolescence? Thefield is in the midst of an explosion of learning aboutthe processes through which neural pathways becomeestablished and influence later behavior, and greaterconsideration is being given to why, from an evolu-tionary perspective, individuals would be particularlyinfluenced by early attachment experiences yet at thesame time retain the ability to flexibly adapt to new cir-cumstances (see Belsky, 1999; Cassidy, in press;George & Solomon, 1999; Siegel, 1999). Emergingdata from studies using the Adult Attachment Inter-view suggest that when adult representations are notcongruent with early attachment experiences (e.g.,when an adult with negative childhood experiences hascurrent attachment-related representations that are se-cure), it is adult representation rather than childhoodexperience that is more closely linked to the attach-ment quality of offspring (and presumably to parentingbehavior; Pearson, Cohn, Cowan, & Cowan, 1994).
A Contextual Perspective
Rather than considering simply the parent’s (pastand present) contribution to the caregiving system, re-cent theorizing has used a complex, contextual ap-proach, in which the parent’s own characteristics areconsidered to play an important role yet are not viewedas solely causal. Parental characteristics are thought tooperate within a larger context that includes character-istics of the child; the larger family system, includingthe marriage, the family size and structure, siblingcharacteristics and relations, and the larger kinship net-work; and the environment, including the safety of theneighborhood, availability of resources, and the socialnetwork. Experimental research with primates has sug-gested that mothering does in fact differ as a functionof these contextual factors (Suomi, 1999). (For furtherdiscussion of a contextual perspective on caregiving,see Belsky & Isabella, 1988; Berlin & Cassidy, 1999,in press; Cowan, Cohn, Cowan, & Pearson, 1996; andGeorge & Solomon, 1999.)
Mothers and Fathers
Contributing additional complexity is the fact thatsome caregivers are mothers and some are fathers;within the contemporary evolutionary perspective, theexistence of separate maternal and paternal caregivingsystems in readily understood. Both mothers and fa-thers are concerned with their own reproductive fit-
89
COMMENTARIES
ness. Yet, because mothers and fathers may differsubstantially in the extent to which the survival of anyone child enhances this fitness, their parenting behav-ior may differ. Compared to fathers, mothers havemore to gain in terms of reproductive fitness from eachchild for several reasons (e.g., mothers’ certainty aboutparental status, shorter reproductive life span, longerinterchild intervals, and greater energy expenditure perchild [during pregnancy and lactation]; see Trivers,1972). Despite these differences, there is almost surelyconsiderable overlap between maternal and paternalcaregiving systems as well.
A Life-Span Perspective
Understanding caregiving would be complex evenif all parents were 25 years old and all children were in-fants. Yet both children and parents develop. Thelife-span perspective characteristic of attachment the-ory leads to consideration of two key questions thatadd further layers of complexity to attempts to under-stand caregiving. First, when does an adult become aparent? Becoming a parent at age 22 has advantagesand disadvantages compared with becoming a parentat age 42; becoming a parent at age 14 brings still an-other set of considerations. Second, what age is thechild who is being parented? Opportunities and chal-lenges—and a parent’s own strengths and weak-nesses—differ when the child is an infant, a toddler, anadolescent.
Summary
Despite increasing theoretical attention to thecaregiving system, there is as yet relatively little re-search related to this theorizing. Much research re-mains to be conducted, and the next decade shouldprove an exciting time on this front. Bridges acrossmultiple theoretical perspectives may serve to furtherenrich our understanding of the caregiving system. Theperspective from which Bell and Richard approachcaregiving differs substantially from that of most con-temporary attachment theorists, who instead aregrounded within a framework of contemporary evolu-tionary biology. It is important to note, before closing,that Bell and Richard’s assessment of attachment the-ory leads them to argue not only that there are scientificflaws in the theory’s proposals about caregiving, butalso that the theory is not particularly sensitive to thechallenges of parenting: “Attachment theory has comeclose to trivializing the difficult conflicts and choicesthat caregivers experience” (this issue). In his clinicalwritings, however, Bowlby (1979, 1988) uniformlyexpressed considerable empathy and compassion forthe difficulties and complex considerations that par-
ents face as they struggle to do the best possible job ofraising their children. Finally, it is important to notethat attachment theory has guided research examiningsome of the questions that have been most central toparents: How are children’s hospital stays best han-dled? How will maternal employment influence chil-dren? Will parents spoil children by responding to theircries? Can children become attached to day care pro-viders, and what qualities in the care provider contrib-ute to the formation of a secure attachment?
Notes
This article was written with support from grants bythe National Institute of Mental Health(RO1MH50773 and RO1MH58907) and the NationalInstitute for Child Health and Development(RO1HD36635). I am grateful to Inge Bretherton,Mary Main, Robert Marvin, and Susan Woodhouse fortheir thoughtful comments on a previous draft of thisarticle.
Jude Cassidy, Department of Psychology, Univer-sity of Maryland, Room 2123L, Biology–PsychologyBuilding, College Park, MD 20742–4411. E-mail:[email protected]
References
Belsky, J. (1999). Modern evolutionary theory and patterns of attach-ment. In J. Cassidy & P. R. Shaver (Eds.),Handbook of attach-ment: Theory, research, and clinical applications(pp.141–161). New York: Guilford.
Belsky, J., & Isabella, R. (1988). Maternal, infant, and social contex-tual determinants of attachment security. In J. Belsky & T.Nezworski (Eds.),Clinical implications of attachment(pp.41–94). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Berlin, L. J., & Cassidy, J. (1999). Relations among relationships:Contributions from attachment theory and research. In J.Cassidy & P. R. Shaver (Eds.),Handbook of attachment: The-ory, research, and clinical applications(pp. 688–712). NewYork: Guilford.
Berlin, L. J., & Cassidy, J. (in press). Understanding parenting: Contribu-tions of attachment theory and research. In J. Osofsky & H. Fitzger-ald (Eds.),Handbook of infant mental health. New York: Wiley.
Bowlby, J. (1956). The growth of independence in the young child.Royal Society of Health Journal, 76,587–591.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic Books.
Bowlby, J. (1973).Attachment and loss: Vol. 2. Separation. NewYork: Basic Books.
Bowlby, J. (1979).The making and breaking of affectional bonds.London: Tavistock.
Bowlby, J. (1984). Caring for the young: Influences on development.In R. Cohen, S. H. Weissman, & B. J. Cohler (Eds.),Parenthood(pp. 269–284). New York: Guilford.
Bowlby, J. (1988).A secure base. New York: Basic Books.Bretherton, I., Biringin, Z., Ridgeway, D., Maslin, D., & Sherman,
M. (1989). Attachment: The parental perspective.Infant MentalHealth Journal, 10,203–221.
90
COMMENTARIES
Bretherton, I., Golby, B., & Cho, E. (1997). Attachment and the tran-sition of values. In J. Grusec & L. Kucszynski (Eds.),Handbookseries: Parenting and children’s internalization of values. NewYork: Wiley.
Bretherton, I., & Munholland, K. A. (1999). Internal working modelsin attachment relationships: A construct revisited. In J. Cassidy& P. R. Shaver (Eds.),Handbook of attachment: Theory, re-search, and clinical applications(pp. 89–114). New York:Guilford.
Cassidy, J. (1999). The nature of the child’s ties. In J. Cassidy & P. R.Shaver (Eds.),Handbook of attachment: Theory, research, andclinical applications(pp. 3–20). New York: Guilford.
Cassidy, J. (in press). Adult romantic attachments: A developmen-tal perspective on individual differences.Journal of GeneralPsychology.
Cowan, P. A., Cohn, D. A., Cowan, C. P., & Pearson, J. L. (1996).Parents’ attachment histories and children’s externalizing andinternalizing behavior: Exploring family systems models oflinkage.Journal of Consulting and Clinical Psychology, 64,53–63.
Fraiberg, S. (1980).Clinical studies in infant mental health: The firstyear of life. New York: Basic Books.
George, C., & Solomon, J. (1996). Representational models of rela-tionships: Links between caregiving and attachment.InfantMental Health Journal, 17,198–216.
George, C., & Solomon, J. (1999). Attachment and caregiving: Thecaregiving behavioral system. In J. Cassidy & P. R. Shaver(Eds.),Handbook of attachment: Theory, research, and clinicalapplications(pp. 649–670). New York: Guilford.
Heard, D., & Lake, B. (1997).The challenge of attachment forcaregiving. London: Routledge.
Kobassa, S. C. (1979). Stressful life events, personality, and health:An inquiry into hardiness.Journal of Personality and SocialPsychology, 37,1–11.
Main, M. (1973).Exploration, play, and cognitive functioning as re-lated to child–mother attachment. Unpublished doctoral disser-tation, Johns Hopkins University, Baltimore, MD.
Main, M. (1981). Avoidance in the service of attachment: A workingpaper. In K. Immelmann, G. Barlow, L. Petrinovich, & M. Main(Eds.),Behavioral development: The Bielefeld Interdisciplin-ary Project. New York: Cambridge University Press.
Main, M. (1990). Cross-cultural studies of attachment organization:Recent studies, changing methodologies, and the concept ofconditional strategies.Human Development, 33,48–61.
Main, M., & Goldwyn, R. (1998).Adult attachment scoring and clas-sification system. Unpublished manuscript, University ofCaliformia at Berkeley.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, child-hood, and adulthood: A move to the level of representation. In I.
Bretherton & E. Waters (Eds.), Growing points in attachment the-ory and research (pp. 66–104).Monographs of the Society for Re-search in Child Development, 50(1-2, Serial No. 209).
Pearson, J., Cohn, D., Cowan, P., & Cowan, C. P. (1994). Earned andcontinuous security in adult attachment: Relation to depressivesymptomology and parenting style.Development andPsychopathology, 6,359–373.
Piaget, J. (1954).The construction of reality in the child. New York:Basic Books.
Pianta, R. C., Marvin, R. S., Britner, P., & Borowitz, K. (1996).Mothers’ resolution of their children’s diagnosis: Organizedpatterns of caregiving representations.Infant Mental HealthJournal, 17,239–256.
Pietromonaco, P. R., & Carnelley, K. B. (1994). Gender and workingmodels of attachment: Consequences for perceptions of self andsocial relationships.Personal Relationships, 1,63–82.
Siegel, D. J. (1999).The developing mind: Toward a neurobiology ofinterpersonal experience. New York: Guilford.
Simpson, J. A. (1999). Attachment theory in modern evolutionaryperspective. In J. Cassidy & P. R. Shaver (Eds.),Handbook ofattachment: Theory, research, and clinical applications(pp.115–140). New York: Guilford.
Slade, A., Belsky, J., Aber, L., & Phelps, J. (1999). Maternal repre-sentations of their toddlers: Links to adult attachment and ob-served mothering.Developmental Psychology, 35,611–619.
Solomon, J., & George, C. (1996). Defining the caregiving system:Toward a theory of caregiving.Infant Mental Health Journal,17,183–197.
Sroufe, L. A., & Fleeson, J. (1986). Attachment and the constructionof relationships. In W. Hartup & Z. Rubin (Eds.),The nature anddevelopment of relationships(pp. 51–71). Hillsdale, NJ: Law-rence Erlbaum Associates, Inc.
Suomi, S. J. (1999). Attachment in rhesus monkeys. In J. Cassidy &P. R. Shaver (Eds.),Handbook of attachment: Theory, re-search, and clinical applications(pp. 181–197). New York:Guilford.
Trivers, R. L. (1972). Parental investment and sexual selection. In B.Campbell (Ed.),Sexual selection and the descent of man,1871–1971(pp. 136–179). Chicago: Aldine-Atherton.
Trivers, R. L. (1974). Parent–offspring conflict.American Zoologist,14,249–264.
Zeanah, C., Benoit, D., Hirschberg, L., Barton, M. L., & Regan, C.(1994). Mothers’ representations of their infants are concordantwith infant attachment classification.Developmental Issues inPsychiatry and Psychology, 1,1–14.
Zeifman, D., & Hazan, C. (1997). Attachment: The bond inpair-bonds. In J. A. Simpson & D. T. Kendrick (Eds.),Evolu-tionary social psychology(pp. 237–263). Mahwah, NJ: Law-rence Erlbaum Associates, Inc.
Parent–Child Emotional Bonds: Loving or Caring?
Patricia Noller and Judith A. FeeneySchool of Psychology
University of Queensland, Australia
The key proposition of Bell and Richard’sthought-provoking article is that attachment theoristshave tended to focus on the needs and motivations un-derlying infants’ attachment behavior, rather than onthe motivation of the adults who nurture them. From
the perspective of these authors, attachment theoryexplains infant attachment behavior in terms of the in-fant’s needs for protection and felt security, but it sayslittle about why parents (particularly quality care-givers) respond in ways that increase the security of the
91
COMMENTARIES
child. We agree that this aspect of the development ofthe parent–child bond (i.e., the motivation of parentsfor caregiving) has tended to be downplayed by attach-ment theorists.
Bell and Richard expound a “connection theoreticalorientation” to the infant–caregiver relationship, with afocus on the emotion of caring as the primary motiva-tion for nurturing behavior toward infants (ordepend-ents,to use their term). Rather than arguing against thebasic principles of attachment theory, their approachseeks to expand the theory to incorporate this new per-spective. In other words, as the authors acknowledge,“the ideas presented here are not an alternative to theattachment theoretical orientation. In fact, they arecompatible with most of the logic that Bowlby and hissuccessors have articulated” (this issue).
Emotion and Caregiving
In our response, we would first like to acknowl-edge several important strengths of Bell and Rich-ard’s theoretical approach. We agree that attachmenttheorists have not paid enough attention to the impor-tance of emotion in parent–child relationships, espe-cially from the perspective of the parent. Emotion isclearly relevant as a motivation for caregiving, aswell as being central to the maintenance of par-ent–child relationships. In particular, Bell and Rich-ard’s emphasis on positive emotion provides a bal-ance to attachment theorists’ primary focus onnegative emotions such as fear, anxiety, and distress.In addition, we agree with the authors that becauseparents have the monopoly on resources in the par-ent–child relationship, it is particularly important tofocus on understanding what motivates them to carefor their children and what factors may impede qual-ity caregiving.
What Emotion Underlies ParentalCaregiving?
Bell and Richard suggest that attachment theoryfails to provide a motivation for parental caregiving.Specifically, they note that, from the perspective ofcurrent evolutionary theory, the protection of children(and hence the survival of the child’s genetic inheri-tance) is an unintended consequence of the existenceof the attachment and caregiving behavioral controlsystems, rather than the goal of these systems. In otherwords, evolutionary processes cannot explain whythese control systems operate. According to the au-thors, the dependent’s attachment behavior can be ex-plained in terms of the desires for security andprotection. They go on to argue that these desires areconsistent with social scientists’ prevailing emphasis
on self-interest as a motivating force, but that parents’nurturing behavior is less easily explained. That is, ifself-interest were the only principle operating, veryfew people would be prepared to take on the demand-ing role of parent.
Based on these considerations, the authors proposethe “emotion” of caring as the primary motivation forparental caregiving. They regard the dyadic emotion ofcaring as an autonomous motivation that can success-fully compete with self-interest; this emotion is ori-ented toward the needs of the dependent, rather thanthe pleasure of the caregiver. This approach to par-ent–child relationships grounds caregiving in emotion,rather than cognition, as highlighted by the authors’vivid description of the parent’s first moments with thenewborn child.
Although we agree that emotion deserves a morecentral place in our understanding of parent–child re-lationships, we have several concerns with this per-spective. First, we are not convinced that attachmenttheory explains parental responses in terms of cogni-tion, to the exclusion of emotion. Bell and Richardacknowledge that Bowlby (1969, 1973, 1980) re-garded attachment relationships as the source of ourmost intense emotions; however, they argue that at-tachment theorists perceive these emotions as the out-come of the attachment bond, rather than its cause.From our perspective, however, it is impossible toseparate the attachment bond from the emotions asso-ciated with it.
Second, we are not convinced that care is the rele-vant emotion, or even that it is an emotion. We wouldtend to argue that the primary emotion is love, and thatcaring is the behavioral expression of that love. Belland Richard’s article gives us the impression that theychoose the termcarebecause of their concerns aboutthe diverse connotations of the wordlove in Westernculture. We have some sympathy with this point ofview, and we recognize that their choice of terms mayhave more to do with semantics than with a particulartheoretical position.
Shaver, Morgan, and Wu (1996), however, pre-sented compelling arguments for conceptualizinglove as an emotion. They suggest that it is importantto distinguish between two forms of love: momentarysurges of affection (similar to joy) and enduring rela-tional love. Momentary surges of affection, in partic-ular, appear to meet the criteria for a basic emotion.In the case of parent–child relationships, we wouldargue that most parents experience enduring rela-tional love for their offspring, but that they also havemomentary surges of love and affection (e.g., whenthey look with adoration on their sleeping child). Inother words, we would argue that parental love is thecritical emotion and that it produces caring behaviors.This formulation is consistent with several theoreticalmodels of love (e.g., Buss, 1988; Noller, 1996;
92
COMMENTARIES
Shaver & Hazan, 1988) that describe behavioral ten-dencies associated with love (such as doing things forthe other).
How Does the Emotion UnderlyingCaregiving Develop?
We also have some doubts as to whether Bell andRichard’s account of parental caregiving explains howthe emotion underlying caregiving behavior develops.The authors place emotion as the central issue incaregiving and attachment relationships, arguing thatemotion is causal, rather than simply a signal of thestate of the behavioral control system. Caring is seen asderiving from the caregiver’s relational feelings to-ward the particular dependent, rather than as a constantneed within the caregiver. As we noted earlier, how-ever, it does not seem feasible to try to distinguish be-tween the attachment bond and the associatedemotions. Hence, it is virtually impossible to establishthe extent to which an emotion such as love (or care)causes caring behavior and the extent to which caringbehavior produces the emotion.
Bell and Richard claim that attachment theory failsto explain why attachment bonds are lasting and irre-placeable. However, we are not convinced that the ex-planation offered by these authors adds anything tocurrent thinking about this issue. It seems tautologicalto say that “the caregiving bond endures because it isthe enduring emotion of caring” (this issue). To fullyunderstand parental caregiving and its motivation, itmay be necessary to go further back in a causal chainand to propose parental investment as a cause of theemotion underlying caregiving.
Evolutionary psychologists such as Buss seem toargue, for example, that parental investment in the off-spring plays a causal role in the evolutionary process,with the potential to provide a motivation for caring.Specifically, Buss (1988) argued that adults’ needs forintimacy (including sexual relationships) and their ten-dency to develop caring and supportive attitudes to-ward sexual partners and their shared offspring can bebest understood from an evolutionary perspective.
On a more specific point, it is also interesting tonote that Bell and Richard, in describing the develop-ment of the emotion of caring, seem to imply that theparent’s connection with the child commences with theparent’s viewing of the child at birth and the child’slooking at the parent. There appear to be some short-comings to this approach. For example, in the earlyweeks of life, babies are not very responsive, with eyemovements being not yet under voluntary control (Pe-terson, 1989). On the other hand, for most parents (al-though perhaps especially mothers), the feeling ofconnection and attachment to the child begins duringpregnancy (Condon, 1993). And although it may be
rather simplistic to see women’s nurturing behavior totheir own infants as “natural,” such behavior probablydoes result, at least in part, from a whole range of natu-ral processes, some of which involve hormonalchanges. In short, parental love (or care) is likely to de-velop through a series of complex processes that beginlong before birth. Similarly, Buss (1988) suggestedthat both proximal causes (e.g., culture, physiology)and ultimate causes (related to our evolutionary past)are needed to fully explain acts of love.
Conceptualizing Quality Caregiving
Bell and Richard criticize attachment theory for itsfailure to explain how responsiveness and sensitivityare produced and why they are important from the par-ent’s perspective. They also argue that attachment the-orists portray parental responsiveness as mechanicaland reactive and as operating like an on-off switch inresponse to the attachment behaviors of the dependent.Although it is true that schematic representations of at-tachment as a behavioral control system tend to implythis mode of operation, it is difficult to reconcile thisview with Bowlby’s recognition (noted earlier), thatattachment bonds are intensely emotional in nature.Perhaps the problem is that any diagrammatic repre-sentation of complex processes is likely to oversim-plify those processes to some extent. Further, the ideathat parents are purely reactive in terms of theircaregiving (providing care only in direct response toattachment behaviors from the dependent) does not fitwith our understanding of attachment theory. Rather,the theory suggests that quality caregivers are “avail-able” to the child; they monitor the child’s needs andare actively involved in meshing their own needs andbehaviors with those of the child (e.g., Isabella &Belsky, 1991).
The authors propose that quality caregiving can bebetter understood in terms of empathy (emotional in-tention to know the other) and responsibility (emo-tional intention to help meet the other’s needs), whichthey see as two direct results of the caregiver’s caringfor the dependent. They acknowledge strong similari-ties between sensitivity and empathy and between re-sponsiveness and responsibility. However, they seeresponsibility as much more proactive than respon-siveness, and they argue that the former concept alsohas the advantage of explaining how the competenceand autonomy of the child are promoted.
Again, we see some limitations to these arguments.First, as noted above, we are not convinced that attach-ment theorists see parental responsiveness as totallyreactive. Second, we find it rather tautological to saythat “within the caregiving process of the connectiontheoretical orientation, [empathy and responsibility]are twodirect results[italics added] of the caregiver’s
93
COMMENTARIES
caring for the dependent” (this issue). How do thesecharacteristics emerge from the caregiving process?How can quality caregiving occur if these characteris-tics are not present from the beginning? These ques-tions remain unanswered. Further, although empathyand responsibility may be important aspects of paren-tal caregiving, these concepts do not help to clarify theorigins of specific forms of problematic caregiving,such as compulsive caregiving, parental overindul-gence, and parental overprotection.
Summary
In summary, Bell and Richard suggest that currentformulations of attachment theory pay little attentionto the emotions and motivations involved in parentalcaregiving and are limited in their potential to explainvariability in caregiving. We agree that emotion isclearly relevant as a motivation for caregiving and thatparents do face important choices and conflicts in theircaregiving role. We also see value in Bell and Rich-ard’s emphasis on positive emotion rather than on neg-ative emotions such as anxiety and distress, which areoften the focus of attachment theorists. In addition, weacknowledge the importance of focusing on under-standing the behavior of the caregiver as well as that ofthe child. At the same time, we believe that some oftheir criticisms of attachment theory are somewhatoverstated and that some of their proposed explana-tions verge on the tautological. Nevertheless, we be-lieve that further refinements of their connection
theoretical orientation are likely to add to our under-standing of the bond between parent and child.
Note
Patricia Noller and Judith A. Feeney, School of Psychol-ogy, University of Queensland, Brisbane, Queensland,4072, Australia. E-mail: [email protected].
References
Bowlby, J. (1969).Attachment and loss: Vol. 1. Attachment. NewYork: Basic Books.
Bowlby, J. (1973).Attachment and loss: Vol. 2. Separation: Anxietyand anger. New York: Basic Books.
Bowlby, J. (1980).Attachment and loss: Vol. 3. Loss. New York: Ba-sic Books.
Buss, D. M. (1988). Love acts: The evolutionary biology of love. InR. J. Sternberg & M. L. Barnes (Eds.),The psychology of love(pp. 100–118). New Haven, CT: Yale University Press.
Condon, J. T. (1993). The assessment of antenatal emotional attach-ment: Development of a questionnaire instrument.British Jour-nal of Medical Psychology, 66,167–183.
Isabella, R. A., & Belsky, J. (1991). Interactional synchrony and theorigins of infant–mother attachment: A replication study.ChildDevelopment, 62,373–384.
Noller, P. (1996). What is this thing called love? Defining the lovethat supports marriage and family.Personal Relationships, 3,97–115.
Peterson, C. (1989).Looking forward through the life span.Sydney,Australia: Prentice-Hall.
Shaver, P. R., & Hazan, C. (1988). A biased overview of the study oflove.Journal of Social and Personal Relationships, 5,473–501.
Shaver, P. R., Morgan, H. J., & Wu, S. (1996). Is love a “basic” emo-tion?Personal Relationships, 3,81–96.
Understanding What Motivates Sensitive Parenting
Theodore DixDepartment of Human EcologyUniversity of Texas at Austin
Fundamental to harmonious and beneficial interac-tions between parents and children is the parents’ moti-vation to promote children’s wants and needs. Byseeking to promote children’s interests, parents mini-mize parent–child conflict; meet children’s legitimateneeds; develop parent–child relationships; and pro-mote children’s sense of trust, competence, andwell-being (Ainsworth, Blehar, Waters, & Wall, 1978;Dix, 1992; Lamb & Easterbrooks, 1981; Tronick,1982). Yet, acting for children’s interests is seldom ac-complished without cost. To read to or comfort chil-dren, to help them resolve conflicts, or to help themfind and play with toys requires effort and means that
at such moments parents cannot work, relax, or attendto others. How do parents decide at each momentwhich outcomes to seek and whether children’s inter-ests should have priority over parents’ or others’ legiti-mate interests? Although important aspects ofchildren’s development may depend on how parents’make these choices, the motivational processes re-sponsible for them have rarely been studied.
Several recent analyses, however, give a centralrole to the processes that regulate which concerns ormotives parents seek to promote with children (Dix,1991, 1992; Hastings & Grusec, 1998). These analysessuggest that, to be effective, parents must be motivated
94
COMMENTARIES
for particular kinds of outcomes, specifically, for out-comes (a) that handle children’s fundamental needs;(b) that children actively seek themselves; (c) that chil-dren are able and willing to promote; (d) that, if pro-moted, would benefit children; and (e) that balance themultiple wants and needs of individuals in the family(Dix, 1991, 1992; see also Kochanska, 1997;Rocissano, Slade, & Lynch, 1987; Tronick, 1982).How and why do effective parents choose to seek out-comes that over time benefit children rather than thosethat may directly benefit parents themselves?
Connection Theory and theMotivational Basis of Parenting
Bell and Richard’sconnection theoryconfronts thismotivational issue by positing an emotion, caring, thatis thought to motivate sensitive parenting, by designat-ing the goal of meeting the child’s needs as the under-lying motivational principle directing sensitiveparenting, and by distinguishing several componentsof the processes that regulate sensitivity, notably car-ing, empathy, and responsibility. This position contrib-utes to analysis of parental competence, first, becauseit recognizes that parents’ emotions are principal, di-recting, and motivating forces behind sensitivity tochildren. Traditional models of parenting, with theiremphasis on learning principles or stable childrearingvalues and attitudes, have been slow to integrate ad-vances in emotion research into analyses of parent be-havior. That parenting must arise “from anautonomous motivation that can successfully competewith self- interest” (this issue) is an important insightthat has only recently influenced analyses of par-ent–child interaction (Dix, 1991, 1992; Hastings &Grusec, 1998).
Second, the connection position makes useful dis-tinctions among components of the regulatory pro-cesses underlying parenting. Notable among these isresponsibility,“the caregiver’s intention to help thepartner meet his or her needs” (this issue). Althoughpresented as an individual difference variable, thisconcept alludes to processes that have been ignored byparenting researchers but that have been important inaction control theory and in social psychology gener-ally (e.g., Ajzen, 1985; Gollwitzer, 1999; Kuhl &Beckmann, 1985). The concept implies that actiondoes not follow directly from appraisals of or emo-tional reactions to events but must be organized by im-mediate situation-specific processes. Sensitiveparenting does not emerge automatically from caringor from empathic cognitions and emotions; rather, itrequires executive processes, that is, processes thattranslate appraisals and emotional tendencies into situ-ation-specific behavioral intentions and action plans(Dix & Gershoff, in press). Even parents who are car-
ing and empathic may act insensitively if the sensitivebehavior that may normally flow from their caring orchild-oriented affect is disrupted by arousal, lack ofconfidence in their ability to control an interaction(Bugental, 1992), or difficulty discerning actions thatmay handle the child’s needs. Knowledge of the pro-cesses that generate parents’ behavioral intentions andaction plans and of the factors that facilitate and under-mine these processes is very much needed.
An Alternate View of the MotivationalBasis of Parenting
Several aspects of connection theory, however, needclarification and development. First, although it givesneeded attention to positive, child-oriented emotions asmotivators of sensitive responses to children, its analy-sisof theseaffectiveprocesses isasyetundevelopedandat times does not mesh well with recent advances inknowledge of emotions.Caring,for example, is said tobe “an enduring dyadic emotion that continues over thelong term” (this issue). By suggesting that caring emo-tions are enduring over the long term, the connectionview implies that caring does not possess the propertiesthat define emotion for most emotion researchers. Thatis, at any moment a parent who is high in caring need not(a) be aroused, (b) have phenomenological experiencesof emotion, (c) have strong motivational inclinations,(d) have emotion-linked behavioral tendencies, or (e)display expressive behaviors (e.g., Campos, Barrett,Lamb, Goldsmith, & Stenberg, 1983; Frijda, 1986;Izard, 1991; Lazarus, 1991; Mandler, 1984). Can caringemotion be present when my daughter is neither presentnor inmythoughts?Cancaringemotionbepresentat thesame moment that I am angry at her? Such questions re-veal the difficulty of thinking of caring both as enduringand relatively unchanging and yet as an emotion.
Guided by recent advances in emotion research, analternate view depicts sensitivity as the result not of anenduring emotion caring, but of a complex set of moti-vational predispositions that regulates how parents setpriorities, interpret events, and react emotionally to thechild (Dix, 1991, 1992). In this view, caring orchild-oriented emotion arises in part because parentsare predisposed to value children’s well-being highly,relative to other outcomes, and therefore to process in-formation that is relevant to children’s well-being, toexperience emotions when children’s well-being issignificantly affected, and often to act to improve chil-dren’s well-being (Dix, 1991, 1992). These child-ori-ented predispositions compete with a multitude ofother concerns, motives, or desires that parents bring tointeractions with children, most of which may have lit-tle to do with children’s interests. At each moment theparents’ problem is to determine which concerns rele-vant to immediate interaction are most important and
95
COMMENTARIES
thus toward which concerns cognition and actionshould be directed. To determine which outcomesshould have priority, parents must continually ap-praise, often automatically and without conscious de-liberation, the interests of children, of parentsthemselves, and of others in the interaction. At eachdecision point, caring or sensitive parents give consid-erable weight to the child’s immediate interests, buteven caring parents often will determine that thechild’s interests are secondary and often will not act topromote those interests. In fact, even quite early in de-velopment, it is critical to socialization that parentshelp children understand and act appropriately whenothers’ wants and needs are more important than chil-dren’s wants and needs.
Within this analysis, the motivational tendenciesthat underlie parental competence are seen to be influ-enced by a complex set of factors that include both sta-ble and unstable aspects of parents, children, andcircumstances. Parents probably differ reliably in theirtendencies to value children’s well-being over otherthings, but at any moment their motivation to promotechildren’s interests depends as well on the parents’mood, immediately preceding events, parents’ apprais-als of the costs of promoting children’s interests, theirappraisal of the probability that particular concerns canbe promoted, and their assessment of children’s moodsand abilities to integrate into particular plans of action.Although over time, sensitive parents ought to experi-ence greater affection, warmth, sympathy, and relatedtransient emotions than will insensitive parents, thesechild-oriented emotions are determined by situationalfactors, characteristics of children, and motivationalpredispositions that need not involvephenomenological affective experience, physiologicalarousal, motivational intensity, immediate behavioralinclinations, and expressive actions. Such emotionalphenomena are the parents’ reactions when attempts topromote important concerns go awry or are fulfilled.Thus, although parents possess dispositions that biasthe likelihood that they will experience particular emo-tions, child-oriented emotions themselves are not en-during and long-term; rather, they occur at momentswhen immediate events undermine or promote the par-ent’s enduring concerns for and interest in the child.
This analysis has several advantages. First, it ac-counts for child-oriented motivation while retaining aconception of emotion that is consistent with modernemotion research, that is, the view that emotions aretransient, organizing, and energizing states that regu-late immediate action plans (e.g., Campos et al., 1983;Dix, 1991; Frijda, 1986; Izard, 1991; Lazarus, 1991;Mandler, 1984). Modern emotion research provides arich foundation for understanding how parents’ emo-tions are activated, what changes emotions bring toparents once aroused, and what processes parents useto regulate them (Dix, 1991). From this vantage point,
deficits in motivation to promote children’s interestscan be understood as motivational-affective processesthat are influenced by stress; support; children’s be-haviors and temperaments; and parents’ skills, atti-tudes, and dispositions. Second, the analysis canaccount not only for variations in sensitivity acrossparents, so often emphasized in parenting research, butalso for variations in sensitivity across situations. Bypositing an ongoing process by which parents discernthe immediate motives or concerns that should havepriority, this analysis can explain why even a caringparent who can be acutely sensitive to a child’s inter-ests on one occasion can be relatively unmotivated bysimilar child interests on another.
This approach yields a somewhat different viewthan that of connection theory of the roles parents andchildren play in parent–child interaction and of themotivational processes underlying sensitive parentbehavior. Although Bell and Richard recognize thatparents have needs and face multiple demands, theiranalysis nonetheless emphasizes that children areprincipally the ones with needs, while parents areprincipally the ones with resources to meet thoseneeds. Often, when children’s needs are pressing andparents’ needs are not, this view captures a criticalaspect of early parent–child functioning. But particu-larly when children’s fundamental needs are rou-tinely met, it is often parents’ needs that are press-ing—they need to complete important work, get to adoctor’s appointment, or attend to an angry neigh-bor—and children’s needs that are not—they simplywant candy, a particular toy rather than others like it,or yet another turn at a game that they have beenplaying for some time. At these moments, it is theparent whose needs are critical; children are simplyparticipants in an interaction in which they must inte-grate into a plan that involves their interests only sec-ondarily. Understanding parent–child interaction re-quires knowledge of how parents assess frommoment to moment the relative importance of multi-ple wants and needs, how they balance the legitimateneeds of children with the legitimate needs of parentsand others, and how they integrate children into on-going plans in which children’s interests are second-ary. Although children’s wishes usually should berecognized, accepted, and, when possible, addressed,often it is not in the best interests of the dyad, thefamily, or even, over the long haul, the child to orientinteraction around children’s immediate needs. Par-ents’ ongoing difficulty reconciling strong demandsfrom stressful environments with strong needs fromtheir children is an important determinant of insensi-tive parenting in many distressed families. The moti-vational, affective, and cognitive processes that par-ents employ to manage complex and changingconcerns that parents, children, and others bring tofamily interaction are as yet poorly understood.
96
COMMENTARIES
Conclusion
The motivational basis of parenting—the processesthat at each moment lead parents to direct action to-ward children’s interests or toward other ends—ispoorly understood. Often researchers assume that dur-ing parent–child interaction parents’ purposes or con-cerns are given or consensual: They seek to respond tochildren’s signals, to meet children’s needs, to get chil-dren to comply, or to encourage particular childrearingvalues. Yet what parents are motivated for with chil-dren, the purposes toward which their behavior is di-rected, vary widely across parents and situations andmay determine how parents appraise and react to chil-dren’s behavior (Dix, 1992; Hastings & Grusec, 1998).In this article I suggest an approach, in many wayscompatible with Bell and Richard’s connection theory,to understanding the processes parents use to set ongo-ing priorities. These poorly understood processes maybe important determinants of how sensitive parents areand thus how well children develop.
Note
Theodore Dix, Department of Human Ecology,University of Texas at Austin, Austin, TX 78712.E-mail: [email protected]
References
Ainsworth, M., Blehar, M. C., Waters, E., & Wall, S. (1978).Pat-terns of attachments.Hillsdale, NJ: Lawrence Erlbaum Asso-ciates, Inc.
Ajzen, I. (1985). From intentions to actions: A theory of planned be-havior. In J. Kuhl & J. Beckmann (Eds.),Action control: Fromcognition to behavior(pp. 11–39). New York: Springer-Verlag.
Bugental, D. B. (1992). Affective and cognitive processes withinthreat-oriented family systems. In I. E. Sigel, A. V.
McGillicuddy-DeLisi, & J. J. Goodnow (Eds.),Parental be-liefs systems: The psychological consequences for children(2nd ed., pp. 219–248). Hillsdale, NJ: Lawrence Erlbaum As-sociates, Inc.
Campos, J. J., Barrett, K. C., Lamb, M. E., Goldsmith, H. H., &Stenberg, R. (1983). Socioemotional development. In P. H.Mussen (Series Ed.), M. Haith, & J. J. Campos (Vol. Eds.),Handbook of child psychology: Vol. 2. Infancy and developmen-tal psychobiology(pp. 783–915). New York: Wiley.
Dix, T. (1991). The affective organization of parenting: Adaptiveand maladaptive processes.Psychological Bulletin, 110,3–25.
Dix, T. (1992). Parenting on behalf of the child: Empathic goals in theregulation of responsive parenting. In I. E. Sigel, A. V.McGillicuddy-DeLisi, & J. J. Goodnow (Eds.),Parental beliefsystems: The psychological consequences for children(pp.319–346). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Dix, T., & Gershoff, E. T. (in press). Measuring parent–child rela-tions. In J. Touliatos, B. F. Perlmutter, & G. Holden (Eds.),Sec-ond handbook of family measurement techniques.NewburyPark, CA: Sage.
Frijda, N. (1986).The emotions.Cambridge, England: CambridgeUniversity Press.
Gollwitzer, P. M. (1999). Implementation intentions: Strong effectsof simple plans.American Psychologist, 54,493–503.
Hastings, P. D., & Grusec, J. E. (1998). Parenting goals as organizersof responses to parent–child disagreement.Developmental Psy-chology, 34,465–479.
Izard, C. E. (1991).The psychology of emotion.New York: Plenum.Kuhl, J., & Beckmann, J. (1985).Action control: From cognition to
behavior. New York: Springer-Verlag.Kochanska, G. (1997). Mutually responsive orientation between
mothers and their young children: Implications for early social-ization.Child Development, 68,94–112.
Lamb, M. E., & Easterbrooks, A. (1981). Individual differences inparental sensitivity: Origins, components, and consequences. InM. E. Lamb & L. R. Sherrod (Eds.),Infant social cognition(pp.127–153). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Lazarus, R. S. (1991). Emotion and adaptation. New York: OxfordUniversity Press.
Mandler, G. (1984).Mind and body: The psychology of emotions andstress.New York: Norton.
Rocissano, L., Slade, A., & Lynch, V. (1987). Dyadic synchrony andtoddler compliance.Developmental Psychology, 23,698–704.
Tronick, E. (1982).Social interchange in infancy: Affect, cognition,and communication.Baltimore: University Park Press.
Motivation for Caregiving From an Ethological Perspective
Mary DozierDepartment of Psychology
University of Delaware
Bell and Richard are accurate in saying that attach-ment theorists have given less attention to the care-giver’s motivation for providing care than to theinfant’s motivation for seeking proximity when dis-tressed. This is not, however, to say that attachmenttheorists have neglected caregiving itself, as impliedby the target article’s title. Caregiving is indeed devel-oped in attachment theory and research. The nature of
the caregiving system as a complement to the infant’sattachment system has been developed, starting withBowlby (1969/1982) and extending to contemporarytheorists (e.g., Solomon & George, 1996); the essentialcomponents of a caregiver’s behavior that lead to dif-ferent attachment organizations of the infant have beencarefully researched (Ainsworth, Blehar, Waters, &Wall, 1978); and the characteristics and life histories of
97
COMMENTARIES
caregivers that lead them to behave in different waysand to have babies with different types of attachmentbehavioral strategies have been carefully described(Main & Goldwyn, in press).
Motivation for Caregiving
The motivation for the baby’s attachment to thecaregiver has seemed of more pressing interest andconcern than the motivation for the caregiver’s provid-ing care. Attachment theory is basically an ethologicaltheory, accounting for attachment behavior in terms ofmaximizing survival of the infant and the infant’sgenes (Bowlby, 1969/1982). The theory suggests thatthe infant was less likely to fall off cliffs, get eaten bypredators, or experience other such calamities if he orshe stayed close to the parent when danger presenteditself. Although such danger may be rarely present inthe same form for the contemporary infant, the systemcontinues to operate in basically the same way: The in-fant is biologically prepared to react to threats by seek-ing out the caregiver. Few have questioned theevolutionary roots of this behavior or the motivationfor the behavior.
But what of the caregiver? Although Bowlby(1969/1982) did not give much attention to the care-giver’s motivation for providing care, others, particu-larly comparative psychologists studyingnonprimates as well as nonhuman primates (Hinde,1982; Levine, Lyons, & Schatzberg, 1999), have de-veloped the motivation for caregiving extensively. In-deed, the issue of motivation for caregiving is appli-cable cross-species. We do not ask why the sea turtlebothers to dig a hole in the sand and buries her eggsthere; we do not ask why the mother chimpanzee al-lows her baby to attach ventrally for months afterbirth; but we ask here why the human mother movesto pick up the crying baby. These various caregivingbehaviors have evolved to maximize the likelihoodthat these caregivers’ genes survive. Each of thesemothers behaves as she does because a number ofsystems support her behaving in this way. Humancaregiving is so closely tied to the biological needs ofthe infant and to the biological drives of the motherthat conscious motivation in the form of cognitive ap-praisal or affect is hardly a necessary concept.
Biological Basis of Caregiving
From a biological perspective, hormones associatedwith pregnancy and lactation appear important in regu-lating caregiving, with oxytocin in particular playing acentral role (Pedersen, 1999). Oxytocin is produced inmuch higher than usual quantities during pregnancyand lactation (Carter, 1998; Carter et al., 1999). Re-
sponsiveness to infants can be facilitated among virginfemale rats by the introduction of increased oxytocin inthe system and can be inhibited by the introduction ofoxytocin antagonists (Pedersen, 1999; Van Leengoed,Kerker, & Swanson, 1987). Higher levels of oxytocinare associated with global changes that may facilitatecaregiving, including a greater calmness and a highertolerance for stress (Carter & Altemus, 1999), moretolerance for monotony (which may accompanycaregiving activities), and greater tolerance for un-usual sensory experiences (Levine, 1983).
At a more specific level relative to caregiving,higher levels of oxytocin are associated with changesin sensitivity to infants’ cues and proclivity to engagein caregiving activities. For example, lactating ratmothers are more sensitive to infants’ cues of distressthan nonlactating females are, although they are lesssensitive to other non–infant-related stressors(Smotherman, Brown, & Levine, 1977). Thus, theirsystems appear designed to respond to infant distress.Further, mothers respond specifically to their own in-fants’ signals of distress. For example, squirrel mon-key mothers have been found to react very specificallyto audiotapes of vocalizations of their own infants butnot to vocalizations of other infants (S. Levine, per-sonal communication, August 18, 1999). Most speciesof primate mothers tend to be disturbed by separationsfrom their infants. For example, rhesus and squirrelmonkey mothers show striking elevations in plasmacortisol when separated from their infants (Coe,Mendoza, Smotherman, & Levine, 1978; Levine,1983; Mendoza, Smotherman, Miner, Kaplan, & Le-vine, 1978).
Bell and Richard suggest that attachment theoryviews the caregiving system as entirely reactive, thatis, operating only in response to infants’ calls. Severaltypes of evidence suggest, however, that the mother isnot only reactive, but proactive as well. For example,when the mother squirrel monkey is separated from theinfant and cannot hear the infant’s vocalizations, shenonetheless shows increases in cortisol (Coe et al.,1978). Also, when reunited with her infant following abrief separation, the squirrel monkey mother shows aquick drop in cortisol although her infant remains dis-turbed. Levine (1983) suggested that this rapid recov-ery by the mother is adaptive in that it helps her toprovide adequate care for the infant, particularly dur-ing times of stress. Thus, caregiving behaviors are notonly reactive to infants’ signals, but proactive in thesense of anticipating needs in the absence of signals.Further, the caregiver often seeks greater proximitywith her young when her infant strays beyond what iscomfortable for her.
Thus, there is strong evidence from the comparativeliterature that the primate mother is evolutionarily pre-pared to provide care for her young. The caregivingand attachment systems have evolved to promote close
98
COMMENTARIES
contact between mother and infant, particularly whenthe infant is threatened.
There are examples of caregivers who provide ade-quate care despite not being biologically prepared toprovide care, however. Among humans, adoptivemothers and foster mothers, for example, are usuallyunrelated biologically to the children for whom theycare. Although there is an increased incidence of abuseand neglect among unrelated dyads (Carter, 1998),many biologically unrelated parents provide nurturingcare that leads to the establishment of secure in-fant–caregiver attachments (Dozier, Albus, Stovall, &Bates, 1999; Singer, Brodzinsky, Ramsay, & Waters,1985). Squirrel monkeys left without their mothers aresometimes “aunted” (Levine, 1983) by nonlactatingfemales in the group. Although biology has not specifi-cally prepared these individuals to care for these in-fants, other systems nonetheless function to supportcaregiving.
The Role of Cognition and Affect inCaregiving
Appraisal of Need
According to Bowlby (1969/1982), the caregiver’sappraisal of the infant’s need for protection is the pri-mary conscious motivator of the caregiver’s behavior.The caregiver is biologically prepared to reestablishcontact with the infant when the infant is threatenedand to maintain proximity even under conditions oflow threat. Thus, it is the assessment of the need forprotection (as codetermined by conditions of proxim-ity; conditions external to the infant, such as presenceof predators or environmental strangeness; and by con-ditions internal to the infant, such as vulnerable health)that motivate caregiving behavior. Bowlby assertedthat appraisal, rather than affect, was primary here. Forexample, he suggested that a distracted mother couldprovide an infant with protection when appraising theneed for protection, even though she did not experi-ence any particular affect relevant to caregiving.
Emotions as Motivating Caregiving
Emotions researchers (e.g., Ekman, 1994; Izard,1992, 1994) hold affect as primary, suggesting that theexperience of affect is key to motivating caregiving, aswell as other behaviors. From the perspective of emo-tions researchers, the caregiver experiences various af-fects, which motivate caregiving behaviors. Forexample, the caregiver may feel distress when the childbecomes frightened, with this distress motivating thecaregiver to provide care for the child. According toemotions theorists, the distracted caregiver described
previously nonetheless experiences some affect, how-ever low in intensity (perhaps including mild distressor even anticipatory guilt), that motivates hercaregiving behavior (C. E. Izard, personal communica-tion, August 16, 1999).
In addition to affect’s serving as a stimulus forcaregiving, it also serves to reward and perpetuatecaregiving behaviors. When caregivers’ behaviors re-sult in the reduction of their own or their infants’ dis-tress, that experience is often rewarding. Morespecifically, caregiving behaviors are negatively rein-forced by the reduction in parents’ own distress (if thatis what is experienced) and in their infants’ distress,and positively reinforced by the pleasant feeling ofhaving had their children respond to their care.
Bell and Richard suggest that attachment theory hasneglected the role of affect in caregiving, and they pro-pose that the primary motivator for caregiving behav-ior is the affect of caring. Bowlby’s position was notthat affect was unimportant to the caregiving system,but only that it was not an explicit motivator of behav-ior. Regardless of the primacy of affect versus ap-praisal, affect, as well as appraisal, are components ofthe caregiving system that assure that most infants areafforded care. I consider Bell and Richard’s proposalthat caring isthe key to motivating caregiving prob-lematic for several reasons. Introducing caring as af-fect stretches the definition of affect beyond what mostemotions scholars are comfortable with (e.g., Ekman,1994; Izard, 1992, 1994). Although caring is surelyemotional, it is much too complex and interpersonal tobe termed an emotion (Izard, 1994). Most important, Ihave argued that caregiving is multiply determined bya host of factors, starting with a maternal biology thatfavors caregiving. To argue that there is any single mo-tivator of caregiving behavior is simplistic.
Notes
The writing of this article was supported by NIMHgrant R01 52135. I gratefully acknowledge the help ofSeymour Levine, Carroll Izard, and Kathleen Albus.
Mary Dozier, Department of Psychology, Univer-sity of Delaware, Newark, DE 19716. E-mail:[email protected]
References
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978).Patterns of attachment: A psychological study of the strange sit-uation. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic.
Carter, C. S. (1998). Neuroendocrine perspectives on social attach-ment and love.Psychoneuroendocrinology, 23,779–818.
99
COMMENTARIES
Carter, C. S., & Altemus, M. (1999). Integrative functions oflactational hormones in social behavior and stress management.In C. S. Carter, I. I. Lederhendler, & B. Kirkpatrick (Eds.),Theintegrative neurobiology of affiliation(pp. 361–371). Cam-bridge, MA: MIT Press.
Carter, C. S., DeVries, A. C., Tayman, S. E., Roberts, R. L., Williams,J. R., & Getz, L. L. (1999). Peptides, steroids, and pair bonding.In C. S. Carter, I. I. Lederhendler, & B. Kirkpatrick (Eds.),Theintegrative neurobiology of affiliation(pp. 169–182). Cam-bridge, MA: MIT Press.
Coe, C. L., Mendoza, S. P., Smotherman, W. P., & Levine, S. (1978).Mother–infant attachment in the squirrel monkey: Adrenal re-sponse to separation.Behavioral Biology, 22,236–263.
Dozier, M., Albus, K. E., Stovall, K. C., & Bates, B. (1999).Maternalstate of mind predicts foster infant attachment.Manuscript inpreparation.
Ekman, P. (1994). Strong evidence for universals in facial expres-sions: A reply to Russell’s mistaken critique.PsychologicalBulletin, 115,268–287.
Hinde, R. A. (1982). Attachment: Some conceptual and biological is-sues. In C. M. Parkes & J. Stevenson-Hinde (Eds.),The place ofattachment in human behavior(pp. 60–76). New York: Basic.
Izard, C. E. (1992). Basic emotions, relations among emotions, andemotion–cognition relations.PsychologicalReview,99,561–565.
Izard, C. E. (1994). Innate and universal facial expressions: Evidencefrom developmental and cross-cultural research.PsychologicalBulletin, 115,288–299.
Levine, S. (1983). A psychobiological approach to the ontogeny ofcoping. In N. Garmezy & M. Rutter (Eds.),Stress, coping, anddevelopment in children (pp. 107–131). New York:McGraw-Hill.
Levine, S., Lyons, D. M., & Schatzberg, A. F. (1999).Psychobiological consequences of social relationships. In C. S.Carter, I. I. Lederhendler, & B. Kirkpatrick (Eds.),The integra-tive neurobiology of affiliation(pp. 83–92). Cambridge, MA:MIT Press.
Main, M., & Goldwyn, R. (in press). Adult attachment rating andclassification system. In M. Main (Ed.),A topology of humanattachment organization assessed in discourse, drawings,and interviews. New York: Cambridge University Press.
Mendoza, S. P., Smotherman, W. P., Miner, M. T., Kaplan, J., & Le-vine, S. (1978). Pituitary–adrenal response to separation inmother and infant squirrel monkey.DevelopmentalPsychobiology, 11,169–175.
Pedersen, C. A. (1999). Oxytocin control of maternal behavior: Reg-ulation by sex steroids and offspring stimuli. In C. S. Carter, I. I.Lederhendler, & B. Kirkpatrick (Eds.),The integrativeneurobiology of affiliation(pp. 301–320). Cambridge, MA:MIT Press.
Singer, L. M., Brodzinsky, D. M., Ramsay, D., Steir, M., & Waters,E. (1985). Mother–infant attachment in adoptive families.ChildDevelopment, 56,1543–1551.
Smotherman, W. P., Brown, C. P., & Levine, S. (1977). Maternal re-sponsiveness following differential pup treatment andmother–pup interactions.Hormone Behavior, 8,242–253.
Solomon, J., & George, C. (1996). Defining the caregiving system:Toward a theory of caregiving.Infant Mental Health Journal,17,183–197.
Van Leengoed, E., Kerker, E., & Swanson, H. H. (1987). Inhibitionof postpartum maternal behavior in the rat by injecting anoxytocin antagonist into the cerebral ventricles.Journal of En-docrinology, 112,275–282.
How (Pro-)Social Is the Caring Motive?
Catrin Finkenauer and Wim MeeusDepartment of Child and Adolescent Studies
Utrecht University
Drawing on their criticisms of attachment theory,Bell and Richard present their connection theory ori-entation, arguing that caregiving is motivated by theemotion of caring. Caring is defined as an enduringemotion that motivates caregivers to meet and gratifythe needs of a specific dependent. It is oriented to-ward the dependent’s needs, rather than those of thecaregiver. The authors emphasize that their theory, incontrast with many theories in contemporary socialscience, “does not consider any self-interest explana-tion adequate to account for caregiving, neither reci-procity nor social norms.… these explanations arelogically unpersuasive … and violate the internal ex-perience of caregiving” (this issue). It may be usefulto shift attention of attachment researchers and theo-rists from one partner in the dyad (i.e., the dependent)to the other (i.e., the caregiver) and to emphasizeother-directed (i.e., prosocial) emotions. However,Bell and Richard may be “pushing the needle a bittoo far.” To enhance our understanding of why par-ents take care of their children (or people of people in
general), we discuss two issues Bell and Richard dis-card from their analysis of the caregiving process.First, we address the issue of self-interest in attach-ment and caregiving. Second, we suggest thatcaregiving has to be considered as a dynamic interac-tion happening between people rather than within orfor one person.
Self-Interest in Caregiving
The most striking and important contribution of thetarget article is its demonstration that the caregivermust be studied in the context of her or his motivationto take care of a specific dependent. Bell and Richardsuggest that this motivation is the entirely other-ori-ented caring of the caregiver for the dependent (whohas “a monopoly on need,” this issue). The caregiver,when holding the dependent for the first time, experi-ences the feeling of caring for the specific dependent.The dependent’s (real or imagined) responsiveness tothe caregiver, that is her or him personally, then initi-
100
COMMENTARIES
ates the caregiving bond (cf. Feeney & Noller, 1996;Reis & Patrick, 1996). Bell and Richard point out that adependent who does not respond or who is perceivedas not responding, as may occur, for example, when thedependent is born blind, may significantly obstruct thecreation of a caring bond. Although the authors restricttheir analysis of the caring bond to the parent–child re-lationship, they suggest that the caregiving logic of theconnection theoretical orientation holds for other so-cial interactions as well (e.g., children caring for aging,tired, or disabled parents). The question arises, how-ever, if the caregiver’s caring is entirely prosocial, whyis it necessary, perhaps even sufficient, that the de-pendent is responsive and “gives something back”?One possible answer is that the caring motive may notbe as other-oriented as Bell and Richard suggest.Rather, it seems necessary to additionally considerself-interest or self-oriented gains.
The literature offers different self-oriented motivesthat may be important in explaining attachment andcaregiving processes. A first motive has its roots in evo-lutionary theories. Although Bell and Richard dismisstheevolutionary functionofcaregivingas inconsequen-tial to the caring process, we contend that a theory oncaregiving that neglects evolutionary functions has toremain incomplete.Bothsurvivalandreproductiveben-efits derive from successful caregiver–dependent inter-actions (e.g., Bowlby, 1969/1982; Baumeister & Leary,1995; Zeifman & Hazan, 1997). Children who succeedin motivating adults to care for them are more likely tosurvive and reach reproductive age. Parents who suc-ceed in providing adequate care for their children in-crease thechance that theiroffspringwill reachmaturityand reproduce in turn. In this sense, a strong and endur-ing mutual bond between caregiver and dependent isclearly adaptive and consequential. The innate qualityof caregiving may have originally evolved to serve evo-lutionary functions, and caring is likely to be partly (butnot necessarily consciously) motivated by evolutionaryself-interest.
Another important motive for caregiving may befound in thehumanneedforbelongingness(Baumeister& Leary, 1995) or relatedness (Reis & Patrick, 1996).People appear to have to an innate drive to form andmaintain at least a minimum quantity of lasting, posi-tive,andsignificant relationships (Baumeister&Leary,1995).Theygotogreat lengths to feel theybelongandtoavoid feeling lonely. In their elegant review,Baumeister and Leary made a convincing argument andgathered a large amount of evidence to show that peopleform relationships easily and readily and that they gen-erally resist losing or breaking off interpersonal rela-tionships, even if these relationships are painful or dis-tressing. Emotional and cognitive patterns show manylinks tobelongingness.Forexample, cognitiveprocess-ing gives priority—in terms of quantity as well as qual-ity—to information that concerns attachment figures
rather than strangers (Sedikides, Olsen, & Reis, 1993).People who have social bonds are better off in a broadvariety of ways (for reviews, see Finkenauer &Baumeister, 1997; Stroebe & Stroebe, 1996). Such peo-ple show a lower mortality rate, less depression, andfewer psychological and physical health problems thanpeople with weak social networks and people who arealone.Thus, there isampleevidence tosuggest thatpeo-ple are motivated to feel belongingness to others andthat theyhavegoodreasons todoso.Conversely,peopleare highly motivated to avoid loneliness, which predis-poses them to mental and physical problems and unhap-piness (Baumeister & Leary, 1995; Reis & Patrick,1996). Enduring social bonds, including caregiving re-lationships, provide people with a feeling ofbelongingness and relatedness, thereby helping them tomaintain their mental and physical well-being. Consis-tent with the suggestion that caregiving relationshipsmay help to satisfy the need to belong is the finding thatone of the major reasons people report for having chil-dren is the fear of being lonely when being old (Blake,1979). Thus, again it appears that caregiving may, inpart,servetomeetself-orientedneedsof thecaregiver.
In contradiction to this suggestion is that fact that theliterature consistently shows that parenthood reducesparents’ happiness and increases feelings of strain andstress (for a review, see Baumeister, 1991). Marriageseems to increase happiness and well-being, but havingchildren seems to considerably decrease these feelings.Despite the price parents seem to pay, parenthood ap-pears to be worth it. Most parents describe that havinghad children represents one of their greatest accom-plishments and report not regretting the experience(Veroff, Douvan, & Kulka, 1981). Also, they generallyfeel that childrenbringparentscloser togetherbygivingthem a shared task and providing shared joys (Hofman& Manis, 1978). In this sense, parenthood appears to bea powerful source of meaning and purpose in the care-giver’s life (Baumeister, 1991).
Taken together, there is abundant evidence to sug-gest that caregivers’ motivation of in caring may not beas prosocial or other-oriented as Bell and Richard pro-pose.The literatureshows thatcaregiving isquitecom-patible with the idea that caring, at least in part, in-volves self-interests, including evolutionary gains,feelings of belongingness, and the provision of mean-ing and purpose.
Caregiving as a Dynamic Interaction
Although Bell and Richard have come to focus onthe extent to which the caregiver cares for the de-pendent, they neglect the fact that the caregiver–de-pendent relationship is interactive and mutuallytransformative. Caring occurs as part of an ongoingsocial interaction between caregiver and dependent,
101
COMMENTARIES
and both interactants contribute to the success of theattachment and caregiving process. Both caregiverand dependent continuously interact to influence eachother (e.g., Bowlby, 1969/1982). Experimental stud-ies have showed not only that parents’ responsivenessinfluences children’s behavior (e.g., Ainsworth,Blehar, Waters, & Walls, 1978), but also that chil-dren’s responsiveness influences parents’ behavior aswell (e.g., Bugental, Caporael, & Shennum, 1980;Cantor & Gelfand, 1977; Schreibman & Koegel,1996). Schreibman and Koegel, for example, report aseries of studies examining the effectiveness of atraining program for parents of autistic children.These children are characterized by severe impair-ments in communicative and social skills, whichplaces enormous strain and stress on the care-giver–dependent relationship. In this program, specialattention is paid to increasing the child’s responsive-ness to (social) cues in the environment. The studiesshowed that increasing the autistic child’s respon-siveness leads parents to spend more pleasant timewith their child. More important, parents themselvesreported being happier and less stressed than parentswho received a treatment that focused on improvingtheir behavior toward their autistic child. The resultssuggest that caring is not entirely unconditional butdepends—in part—on the child’s responsiveness tothe parent. In this sense, caring is an inherently socialphenomenon, in that the caregiver influences the de-pendent, and the dependent, in turn, influences thecaregiver. Although Bell and Richard mention thiscontingency (i.e., parents feel unloved and uncaredfor when children are unresponsive), this aspect ofthe parent–child interaction is not included in theiranalysis of caregivers’ caring. Considering attach-ment and caregiving processes from the caregiver’sperspective does not seem to do justice to the dy-namic and mutual nature of the caring relationship.Conversely, it does not seem adequate to consider at-tachment processes only from the dependent’s per-spective. Rather, the ongoing process of interactionbetween caregiver and dependent needs to be in-cluded in theories on attachment and caregiving.
Conclusion
Although Bell and Richard’s connection theoreti-cal orientation draws attention to important issuessurrounding the dependent–caregiver relationship,their case rests on an inadequate consideration of the(pro-)social aspects of the caregiving relationship.First, the connection theoretical orientation seems tooverrate the prosocial nature of the caring motive.The literature suggests that self-interest motives arepowerful mechanisms in interpersonal relationshipsthat need to be considered to enhance our understand-
ing of attachment and caregiving processes. Second,the connection theoretical orientation seems to under-rate the truly social nature of the caregiving relation-ship. Bell and Richard’s suggestions are limited to anindependent analysis of one partner in the caregivingrelationship, namely the caregiver. More studies andtheories are needed that consider attachment andcaregiving as reciprocal, dynamic, and mutuallytransformative, in that each partner’s behavior,thoughts, and emotions influences the other’s behav-ior, thoughts, and emotions.
Note
Catrin Finkenauer and Wim Meeus, Department ofChild and Adolescent Studies, Utrecht University,3508 TC Utrecht, The Netherlands. E-mail:[email protected] or [email protected]
References
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Walls, S. (1978).Patterns of attachment: Assessed in the strange situation and athome. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Baumeister, R. F. (1991).Meanings of life. New York: Guilford.Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire
for interpersonal attachment as a fundamental human motiva-tion. Psychological Bulletin, 117,497–529.
Blake, J. (1979). Is zero preferred? American attitudes toward child-lessness.Journal of Marriage and the Family, 41,254–257.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic Books.
Bugental, D. B., Caporael, L., & Shennum, W. A. (1980). Experi-mentally produced child uncontrollability: Effects on the po-tency of adult communication patterns.Child Development, 51,520–528.
Cantor, N. L., & Gelfand, D. M. (1977). Effects of responsivenessand sex of children on adults’ behavior.Child Development, 48,232–238.
Feeney, J., & Noller, P. (1996).Adult attachment. Thousand Oaks,CA: Sage.
Finkenauer, C., & Baumeister, R. F. (1997). L’effet des variablessubjectives sur le bonheur: Résultats de la recherche et impli-cations pour la thérapie [The impact of subjective variableson happiness: Empirical findings and their implications forpsychotherapy].Revue Québécoise de Psychologie, 18,99–118.
Hofman, L., & Manis, J. (1978). Influences of children on maritalquality and family interaction. In R. Lerner & G. Spanier (Eds.),Child influences on marital and family interaction: A life-spanperspective(pp. 165–213). New York: Academic.
Reis, H. T., & Patrick, B. C. (1996). Attachment and intimacy: Com-ponent processes. In E. T. Higgins & A. W. Kruglanski (Eds.),Social psychology: Handbook of basic principles(pp.523–563). New York: Guilford.
Schreibman, L., & Koegel, R. L. (1996). Fostering self-manage-ment: Parent-delivered pivotal response training for chil-dren with autistic disorder. In E. D. Hibbs & P. S. Jensen(Eds.), Psychosocial treatments for child and adolescentdisorders: Empirically based strategies for clinical prac-
102
COMMENTARIES
tice(pp. 525–552). Washington, DC: American Psychologi-cal Association.
Sedikides,C.,Olsen,N.,&Reis,H.T. (1993).Relationshipsasnaturalcate-gories.Journal of Personality and Social Psychology, 64,71–82.
Stroebe, W., & Stroebe, M. (1996). The social psychology of social sup-port. InE.T.Higgins&A.W.Kruglanski (Eds.),Socialpsychology:Handbook of basic principles(pp. 597–621). New York: Guilford.
Veroff, J., Douvan, E., & Kulka, R. A. (1981).The inner Ameri-can: A self-portrait from 1957 to 1976. New York: BasicBooks.
Zeifman, D., & Hazan, C. (1997). Attachment: The bond inpair-bonds. In J. A. Simpson & D. Kenrick (Eds.),Evolutionarysocial psychology(pp. 237–263). Mahwah, NJ: LawrenceErlbaum Associates, Inc.
Do We Need to Invent Another Emotion?The Role of Caring in Attachment
Eric YoungstromDepartment of Psychology
Case Western Reserve University
The attachment model initially articulated byBowlby (1973) has proven to be a powerful metaphorfor human behavior. Attachment has captured theimagination of ethologists and psychologists, and ithas informed research in developmentalpsychopathology (e.g., Cicchetti, 1994; Cicchetti &Barnett, 1991; Egeland, Pianta, & Ogawa, 1996) aswell as in normal development (e.g., Egeland &Hiester, 1995). More recent work has focused on theinternal working models and attachment styles ofadults, exploring the way that these correlate withother measures of psychosocial functioning. Bell andRichard’s treatment of attachment is consistent withthis trend: Their conception of attachment pertains tosocial connections generally and not solely to the care-giver–child relationship. In contrast to much previousattachment literature, Bell and Richard concentrate onthe behavior of the caregiver (and not the dependent),developing an active role for the caregiver. They ac-complish this shift in focus and role by introducing theconcept ofcaring as a sustained emotion that moti-vates caregiver empathy, sensitiveness, and responsi-bility. These in turn motivate the caregiver’sattachment behaviors, according to the connection the-oretical orientation.
Bell and Richard are careful to define their terms,using explicit—if somewhat idiosyncratic—defini-tions of each of these constructs. Their treatment of“caring” as an emotion represents a departure from themain schools of thought about emotion. To their credit,the authors acknowledge this (see Footnote 3, Bell &Richards, this issue). They base their decision to cate-gorize caring as an emotion on the view that “episodicemotions do not seem capable of performing as moti-vations for ongoing and enduring action (such asparenting)” (this issue). This position raises several re-lated questions, beginning with, “Iscaringan emotionin any accepted technical sense of the term?” If not,then are there other generally accepted emotion con-
structs that could fill the role assigned to caring in theconnection theoretical model? Finally, how bestshould we conceptualize caring behavior in a way thatis consistent with established constructs in psychol-ogy? These questions form the basis for the rest of thiscommentary.
Is Caring an Emotion?
Bell and Richard acknowledge a large body of emo-tion literature before dismissing it and asserting thatcaring is an emotion. The number of discrete emotionsand the canon of different human emotional experi-ences has been widely debated. More evolutionarilyoriented models of emotion (e.g., Izard, 1972;Plutchik, 1980; Tomkins, 1963) have identified vari-ous discrete emotions. These are posited to haveevolved with different motivational functions and dis-tinct neurophysiological pathways and expressive be-haviors. Other emotion experiences are construed aseither a blend of the more “basic” or primary emotions(Plutchik, 1994) or as involving complex interactionsbetween emotions and cognition (Izard, 1993). In con-trast, cognitive-constructivist models of emotion (Laz-arus, 1991; Mandler, 1990) concentrate on dimensionsof affective experience and the role of cognitive ap-praisal in eliciting emotion activation. Interestingly,caring has not been identified as a discrete emotion,nor has it played a prominent role in constructivistmodels of emotion. This omission is clearly at oddswith the pivotal role that Bell and Richard assign tocaring, placing it at the center of affiliative processes.
The fact that caring has not figured prominently inthe major past approaches to emotion does not neces-sarily exclude caring from being classified as an emo-tion. The number of emotions identified as separateconstructs depends on methodology and level of analy-sis, as well as theoretical allegiance. For example, it
103
COMMENTARIES
has been difficult to identify clear peripheral auto-nomic processes that differentiate between emotions(cf. Ekman, Levenson, & Friesen, 1983; Lang, Green-wald, Bradley, & Hamm, 1993; Schachter & Singer,1962). Factor analytic approaches to self-reports oftrait emotionality generally identify two (or sometimesthree; Russell & Mehrabian, 1977) dimensions of emo-tionality (Feldman, 1995; Tellegen, Watson, & Clark,1999). Approaches that concentrate on facial expres-sions identify a larger number of apparently distinctemotions (Ekman & Friesen, 1971; Izard, 1971), butthis number still pales in comparison with the numberof terms identified in emotional lexicons catalogingthe various terms available within a language (Ortony,Clore, & Foss, 1987). Caring appears only in this last,most molecular level of analysis. This places caring atgreatest remove from evolutionary models of emotiondevelopment, from basic dimensions of affective expe-rience, and from readily identifiable expressive behav-iors that characteristically signal the presence of anunderlying emotion state.
In positing that caring is actually an emotion, al-beit unrecognized by past investigators, it is incum-bent on Bell and Richard to (a) present a definition ofemotion and show that caring meets the requisite cri-teria, and (b) demonstrate that caring diverges fromother accepted emotions. There is enough plurality inthe existing definitions of emotion that caring couldgarner at least some support. Bell and Richard wouldneed to pick a clear allegiance with one of the extanttheories and develop support for the ways in whichcaring is congruent with their chosen definition. Evenat a general level, the description of caring developedin the target article poses some problems. Althoughcurrently there is no consensually accepted definitionof emotion, most researchers agree that emotions aretypically brief, episodic responses (Ekman &Davidson, 1994). The most enduring affective statesare usually construed asmoods(e.g., Thayer, 1996).If caring typically had a long duration, it more readilycould be classified as a mood than an emotion in thisregard. On the other hand, it could be argued that car-ing does meet several other criteria for considerationas an emotion, including involving a “core relationaltheme” (Lazarus, 1991), possessing intentionality inthe philosophical sense that it “takes an object” (Sol-omon, 1993), and influencing action more than cog-nition (Davidson, 1993). Currently these are infer-ences based on Bell and Richard’s presentation, notexplicit points incorporated into the connection theo-retical orientation.
Additionally, Bell and Richard would need to pro-vide evidence that caring diverges from the other iden-tified emotions. Caring should be distinguishable fromother emotion states by virtue of physiological indices,neuroimaging techniques, distinct facial expression,reports of subjective experience that show good inter-
nal consistency yet low correlations with other emo-tion criteria, or other measurable behavioral correlates.Given the complexity of human behavior, it is unlikelythat caring would satisfy all of these criteria (nor doesany other emotion across all published studies). Still, itwould be important to demonstrate at least some ofthese characteristics. This is not an impossible stan-dard: jealousy represents a human emotion withoutidentified modal physical or behavioral signals, but itis clearly a powerful potential motivator of behaviorfrom an evolutionary psychology perspective (Buss,1994; Buunk, Angleitner, Oubaid, & Buss, 1996). Re-cent work has tackled the issues of measuring jealousy,and progress has been made in determining distinctelicitors and behavioral correlates (e.g., Salovey & Ro-din, 1986). Proponents of the connection theoreticalorientation to attachment would need to conduct simi-lar basic research on the measurement of caring. With-out this sort of multimethod, multitrait approach todefining caring as a construct, we cannot be sure thatother emotions may not more parsimoniously explainattachment behaviors for which Bell and Richard in-voke caring.
Are There Other Emotions That CouldFill the Role of Caring?
In the absence of a compelling case for caring as anemotion that is distinct from other generally acceptedemotion states, it is logical to ask whether there areother emotions that could play an important role in at-tachment processes. Attachment necessarily involvespatterns of behavior that are sustained over periods ofdays, months, or even years. Bell and Richard arguethat the durational brevity of the widely recognizedemotion states therefore makes them unsuitable formotivating attachment behaviors.
It is here, when describing the subjective experi-ence of caring, that Bell and Richard succumb to rhap-sodizing about caregiving. The presentation slips into ahighly mythologized view of caregiving, conflatingethnotheories about romantic love, idealizedparenting, and even the moral philosophy of Levinasinto a warm, fuzzy tangle. Unfortunately, the true busi-ness of caregiving is far more prosaic. Not every diaperor feeding is a spiritual encounter, nor do these eventsautomatically cumulate with past experiences todeepen one’s level of caring. An objective observerwould conclude that even the most attentive parentspends large parts of the day engaged in activities unre-lated to caregiving. Similarly, much of caregiving be-havior is produced by habit without any pronouncedaffective content. Functional behavioral assessmentwould indicate that many caregiving actions areoperantly conditioned through negative reinforcement:The caregiver will cycle through a series of behaviors
104
COMMENTARIES
(e.g., try feeding, changing, rocking) until the depend-ent stops emitting an aversive stimulus. Still othercaregiving behaviors, such as protectiveness, involveother distinct emotions, such as anger (Izard, 1991).
In this regard, Bell and Richard’s depiction of caringas a state persisting fairly constantly throughout thecourse of the relational connection appears unrealistic.On the other hand, the brief, episodic emotions dis-missedbyBellandRichardactuallydoappearsufficientto develop and maintain complex patterns of social in-teraction. For example, the complex dominance hierar-chies of humans and many other social mammals aremarked by long periods of homeostasis punctuated byoccasional displays of aggression (Lorenz, 1966).These displays of anger and hostility are the exception,not the norm, and they typically occur in moments ofthreatorwhen individualsareattempting tochangetheirplace in the hierarchy (Wilson, 1980). In like manner,brief emotiondisplaysby thedependent, suchascrying,are sufficient to elicit caregiving behaviors. Episodicemotion events also can organize memories andcognitions, carving channels along which future per-ceptions and behavior will run.
Thus, we do not need to rule out brief emotion reac-tions as potentially contributing significantly to attach-ment transactions. In fact, a variety of differentemotions are likely to be called into the service ofcaregiving as events dictate. At times, caregivers willfeel anger, fear, happiness, interest, and other emotionsin the context of the attachment relationship. No singleemotion appears adequate to perform all the functionsthat Bell and Richard attribute to caring. Perhaps this isbecause caring, as described by Bell and Richard, isnot best considered an emotion.
If Not an Emotion, Then What IsCaring?
Given the plurality of theories of emotion and thelack of research into caring per se, we cannot concludethat caring fails to satisfy some criteria for consider-ation as an emotion. We also cannot be sure that it willfail to show divergent validity from other emotions.However, there clearly are problems with Bell andRichard’s current formulation of caring as an emotion.On the other hand, there are no obvious other candidateemotions that perform the functions that Bell and Rich-ard ascribe to caring. Bell and Richard also make acompelling case for the importance of caring (or some-thing like it) in a model of caregiving behavioral sys-tems. If caring, as described by Bell and Richard, doesnot satisfy formal criteria for being an emotion state,then what is caring?
On one level, it is useful to consider parental care asa biological trait with a substantial amount of geneticprogramming. Parenting behaviors vary widely across
species, evidently because of the sensitivity ofparenting behavior to natural selection processes (Wil-son, 1980). Humans, although they do not show closedinstincts that produce rigidly determined behaviors,are not blank slates whose responses to others will bedetermined entirely by learning history (Stevens,1982). At the same time, although maternal care in hu-mans and primates possesses some basic, innate com-ponents, the behaviors are complicated enough torequire practice (in the form of play, as well as directexperience; Wilson, 1980).
Ethologists came relatively late to the study ofparenting behavior, focusing first on aggression andreproduction. However, Eibl-Eibesfeldt pioneered thestudy of parenting behavior, producing insights re-markably consonant with Bell and Richard’s descrip-tion of caring. Eibl-Eibesfeldt (1974) made the pointthat caregiving and dependent behaviors are flexible inhumans and therefore can be used symbolically. Ac-cording to Eibl-Eibesfeldt (1989), the behaviors of thecaregiver and dependent form the basis for affiliativebehaviors in social animals, playing a prominent rolein love and friendship in humans. Eibl-Eibesfeldt char-acterized the parent–child behavior system as one ofthe fundamental components of social behavior, givingit at least equal weight with aggression (Midgley,1978).
This is fertile ground for Bell and Richard. Drawingon ethological work would be consistent with the ori-gins and growth of attachment theory. Both Bell andRichard and Eibl-Eibesfeldt (1974) argued that the lan-guage of caring is the medium of social connection, notjust a narrow attachment bond. Eibl-Eibesfeldt’s workalso provides an evolutionary context and a catalog ofethological observations that could inform Bell andRichard’s connection theoretical model. Like Bell andRichard, Eibl-Eibesfeldt found that the roots of loveare in the caregiver–dependent relationship, not in sex-uality. According to this view, Freud had things back-ward: The mother is using caresses from the caregivingbehavioral repertoire to woo her mate, not using sexualgestures to soothe her child (Midgley, 1978).
Conclusion
The goals of the target article are ambitious. Belland Richard seek to elaborate the motivating forces forcaregiving behavior, turning the caregiver into a moreactive participant in the dyadic exchange. They alsowant to restore emotion to a central position in attach-ment models, rehumanizing a theory whose discussionof cybernetic systems and feedback loops often soundslike it could be describing guided missiles as well aspeople (Stevens, 1982). Finally, Bell and Richard seekto describe caregiver–dependent behavioral systems asa general component of social connections, not as
105
COMMENTARIES
something limited to a privileged set of attachment re-lationships. These are worthy intentions, and theymatch trends in current attachment research that seekto explore adult attachment styles (George, 1984) andtheir correlates with psychological functioning (e.g.,Dozier, Cue, & Barnett, 1994; Dozier, Stevenson, Lee,& Velligan, 1991).
Bell and Richard attempt to accomplish these ambi-tions by invoking caring as the motivator ofcaregiving-related behaviors and feelings—sensitiv-ity, empathy, and responsibility—and then labelingcaringas an emotion to reclaim emotion’s place in theconnection model. As staged here, this argument is adeus ex machina. The solution is too tidy, and there areunsatisfying loose ends when the audience reflects onthe evidence offered that caring is an emotion. Theconnection theoretical orientation need not founder onthis rock. It is possible that a careful argument could bemade that caring is an emotion. More fruitfully, Belland Richard could return to ethology and integrate thework of Eibl-Eibesfeldt (1989) and colleagues (Keller,Schoelmerich, & Eibl-Eibesfeldt, 1988), who give pri-macy to caregiving as a motivational-behavioral sys-tem. Attention then could be directed to articulatinghow different emotions, cognitions, and behaviors arerecruited in service of the parental care system, as wellas to exploring how the behaviors in this repertoire areused in other social connections.
Notes
Thanks to Jennifer Kogos and Carla KmettDanielson for their comments and discussion.
Eric Youngstrom, Department of Psychology, CaseWestern Reserve University, 11220 Bellflower Road,Cleveland, OH 44106–7123. E-mail: [email protected]
References
Bowlby, J. (1973).Attachment and loss: Vol. 2. Separation: Anxietyand anger.New York: Basic Books.
Buss, D. M. (1994).The evolution of desire: Strategies of human mat-ing. New York: Basic Books.
Buunk, B. P., Angleitner, A., Oubaid, V., & Buss, D. M. (1996). Sexdifferences in jealousy in evolutionary and cultural perspective:Tests from the Netherlands, Germany, and the United States.Psychological Science, 7,359–363.
Cicchetti, D. (1994). Developmental processes in peer relations andpsychopathology.Development and Psychopathology, 7,587–589.
Cicchetti, D., & Barnett, D. (1991). Attachment organization in mal-treated preschoolers.Development and Psychopathology, 3,397–411.
Davidson, R. J. (1993). The neuropsychology of emotion and affec-tive style. In M. Lewis & J. M. Haviland (Eds.),Handbook ofemotions(pp. 143–154). New York: Guilford Press.
Dozier, M., Cue, K. L., & Barnett, L. (1994). Clinicians as caregivers:Role of attachment organization in treatment.Journal of Con-sulting & Clinical Psychology, 62,793–800.
Dozier, M., Stevenson, A. L., Lee, S. W., & Velligan, D. I. (1991).Attachment organization and familial overinvolvement foradults with serious psychopathological disorders.Developmentand Psychopathology, 3,475–489.
Egeland, B., & Hiester, M. (1995). The long-term consequences ofinfant day-care and mother–infant attachment.Child Develop-ment, 66,474–485.
Egeland, B., Pianta, R., & Ogawa, J. (1996). Early behavior prob-lems: Pathways to mental disorders in adolescence.Develop-ment and Psychopathology, 8,735–749.
Eibl-Eibesfeldt, I. (1974).Love and hate: The natural history ofbehavior patterns. (G. Strachan, Trans.). New York:Schocken.
Eibl-Eibesfeldt, I. (1989).Human ethology. Hawthorne, NY: AldineDe Gruyter.
Ekman, P., & Davidson, R. J. (Eds.). (1994).The nature of emo-tion: Fundamental questions. New York: Oxford UniversityPress.
Ekman, P., & Friesen, W. V. (1971). Constants across cultures in theface and emotion.Journal of Personality and Social Psychol-ogy, 17,124–129.
Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomicnervous system activity distinguishes among emotions.Sci-ence, 221,1208–1210.
Feldman, L. A. (1995). Variations in the circumplex structure ofmood. Personality and Social Psychology Bulletin, 21,806–817.
George, C. C. Y. (1984).Individual differences in affective sensitiv-ity: A study of five-year-olds and their parents (empathy, attach-ment).Unpublished doctoral dissertation, University of Califor-nia, Berkeley.
Izard, C. E. (1971).The face of emotion. New York: Appleton-Cen-tury-Crofts.
Izard, C. E. (1972).Patterns of emotions. San Diego, CA: Academic.Izard, C. E. (1991).The psychology of emotions. New York: Plenum.Izard, C. E. (1993). Four systems for emotion activation: Cognitive
and noncognitive processes.Psychological Review, 100,68–90.
Keller, H., Schoelmerich, A., & Eibl-Eibesfeldt, I. (1988). Commu-nication patterns in adult–infant interactions in Western andnon-Western cultures.Journal of Cross-Cultural Psychology,19,427–445.
Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm, A. O.(1993). Looking at pictures: Affective, facial, visceral, and be-havioral reactions.Psychophysiology, 30,261–273.
Lazarus, R. S. (1991).Emotion and adaptation. New York: OxfordUniversity Press.
Lorenz, K. (1966).On aggression. London: Methuen.Mandler, G. (1990). A constructivist theory of emotion. In N. L.
Stein, B. Leventhal, & T. Trabasso (Eds.),Psychological andbiological approaches to emotion(pp. 21–44). Hillsdale, NJ:Lawrence Erlbaum Associates, Inc.
Midgley, M. (1978).Beast & man: The roots of human nature. Lon-don: Methuen.
Ortony, A., Clore, G. L., & Foss, M. A. (1987). The referential struc-ture of the affective lexicon.Cognitive Science, 11,341–364.
Plutchik, R. (1980).Emotion: A psychoevolutionary synthesis. NewYork: Harper & Row.
Plutchik, R. (1994).The psychology and biology of emotion. NewYork: HarperCollins.
Russell, J. A., & Mehrabian, A. (1977). Evidence for a three-factor the-oryofemotions.JournalofResearch inPersonality,11,273–294.
Salovey, P., & Rodin, J. (1986). The differentiation of social-com-parison jealousy and romantic jealousy.Journal of Personalityand Social Psychology, 50,1100–1112.
106
COMMENTARIES
Schachter, S., & Singer, S. (1962). Cognitive, social, and physiologi-cal determinants of emotional state.Psychological Review, 69,379–399.
Solomon, R. C. (1993). The philosophy of emotions. In M. Lewis & J.Haviland (Eds.),Handbook of emotions(pp. 3–15). New York:Guilford.
Stevens, A. (1982).Archetypes: A natural history of the self. NewYork: Quill.
Tellegen, A., Watson, D., & Clark, L. A. (1999). On the dimensional andhierarchicalstructureofaffect.PsychologicalScience,10,297–303.
Thayer, R. E. (1996).The origin of everyday moods: Managing en-ergy, tension, ansd stress. New York: Oxford.
Tomkins, S. S. (1963).Affect, imagery, consciousness: Vol. 2.NewYork: Springer.
Wilson, E. O. (1980).Sociobiology: The Abridged Edition. Cam-bridge, MA: Harvard University Press.
Who Cares? For Whom and When, How, and Why?
Ellen BerscheidDepartment of PsychologyUniversity of Minnesota
W. Andrew CollinsInstitute of Child Development
University of Minnesota
One wishes Bell and Richard every success in theirtheoretical attempt to rescue caregiving from the landof constructs attachment researchers forgot. Lack of at-tention to the caregiving side of the attachment theoret-ical equation has been evident for some time. Of the 36chapters in the recently publishedHandbook of Attach-ment(Cassidy & Shaver, 1999), for example, only oneappears to be directly devoted to caregiving. The ne-glect of caregiving in the very domain in which onewould expect it to receive a full-court press, both theo-retically and empirically, has been puzzling. It hasbeen puzzling not only for scientific reasons but forpractical reasons as well. Popular recognition of theperilous position of many infants and young childrenas a result of inadequate caregiving has increased in re-cent years. Teenage pregnancy, single motherhood,dual-career families, marital dissolution, fragmenta-tion of the extended kinship network, and drug and al-cohol abuse are among the many factors that have beenfrequently implicated by politicians, social commenta-tors, and others for the failure of many children in theUnited States to receive the care they need. Thus it isespecially curious that widespread societal concern hasnot moved the swelling army of attachment research-ers to examine the caregiving construct.
The reason for the neglect of attachment theory’scomplementary behavioral system may be more mun-dane than that offered by Bell and Richard: Bowlby’s(1969/1982, 1988) cursory theoretical elaboration ofthe caregiving system may be less responsible than thebalkanizationofpsychology is.Somepsychologists, al-thoughnotattachment researchers,havedevotedagreatdeal of attention to caregiving under the rubrics of “so-cial support” (e.g., Pierce, Lakey, Sarason, & Sarason,1997), “altruism” (e.g., Batson, 1998), and “prosocial
behavior” (e.g., Clark, 1991). In addition, many re-searchers in both developmental and social psychologyhave learned a great deal aboutempathic accuracy,acompetency associated with accurately detecting an-other’sneedsandwishes (see Ickes,1997)and thuswitheffective caregiving. Bell and Richard neither incorpo-rate, nor even mention, these vast literatures despitetheirstatement that they“have introducedtheconceptofconnection todescribevariableemotionalprocessesoc-curring in the full range of relationships” (this issue).One takes the phrase “full range of relationships” tomean relationships between two adults as well as rela-tionships between an adult and a child. Social psycholo-gists have focused almost exclusively on caregiving be-tween adults and, with only a sprinkling of exceptions(e.g., Carnelley, Pietromonaco, & Jaffe, 1996; Feeney,1996; Kunce & Shaver, 1994; Rholes, Simpson, &Orina, 1999; Simpson, Rholes, & Nelligan, 1992), havenot made contact with attachment theory and researchwhen investigating caregiving. The adult–child andadult–adult caregiving literatures thus have developedindependently of each other and remain isolated; devel-opmental psychologists have focused on the child (andattachment) side of the adult–child relationship,whereas social psychologists have concentrated almostexclusively on caregiving in adult–adult relationships.A theory of caregiving that aspires to account forcaregiving behavior in all relationships must integratethese two bodies of knowledge.
Bell and Richard charge that attachment theory failstogiveanadequateaccountof caregiving inanumberofways. They observe, for example, that “attachment the-ory usually describes the caregiving behavioral controlsystemaspurely reactive” (this issue),activatedbysuchchild attachment behaviors as crying or other distress
107
COMMENTARIES
signals. To remedy this deficiency, they intend theirConnection Model to account for “proactive” as well asreactive caregiving behaviors. This is an especiallylaudable aim, for we frequently see people engaging inanticipatory and unsolicited caregiving behaviors in re-sponse to others’ needs, as the authors discuss. We alsosee people actively seeking out others to care for (notonly other humans, but animals, pet rocks, and, more re-cently, “Furbies”), and this behavior, too, requires ex-planation in any model of caregiving. Whether searchesfor caregiving opportunities have as their aim the wel-fare of potential recipients of care or the welfare of thecaregiver is likely toproveasthornyaproblemtoattach-ment and other caregiving researchers as it has to re-searchers of altruism. Some researchers, however, havemade progress in identifying those persons who arelikely toengage incaregiving forself-enhancementpur-poses (e.g., Helgeson, 1994).
The raison d’être for Bell and Richard’s ConnectionModel is said to be the failure of attachment theory tospecify a proximal motivation for caregiving in theparent–child relationship (the “ultimate” motivationbeing supplied by evolutionary theory, as the authorsnote). The motivation for caregiving posited by theConnection Model is “the emotion of caring.” The au-thors contend that this emotion not only supplies a mo-tivation for caregiving but also helps solve anotherproblem: that “the role of emotion in attachment andcaregiving is still seriously underexplained” (this is-sue). The authors neglect to say, however, that onevery important reason for this state of affairs is thatemotion itself is still “underexplained” in psychology.After a century of effort, heated controversy still per-vades answers to even the most fundamental questionsabout emotion (see Ekman & Davidson, 1994).
It is easy to agree with Bell and Richard that the is-sue of a motivation for caregiving needs attention, notsimply to redress the alleged lacuna in attachment the-ory but for the very practical reasons mentioned ear-lier. But those familiar with contemporary theory andresearch on emotion will not be optimistic that thenewly minted emotion of “caring,” added to the daunt-ingly long list of emotions people have posited over theyears, will prove useful either to our understanding ofcaregiving or to our understanding of emotional expe-rience in relationships in which caregiving and care-re-ceiving behaviors are distinct features of the partners’interaction pattern. At minimum, elaboration of thiscenterpiece of the Connection Model will requirecloser contact with contemporary emotion theory andresearch. For example, in a footnote, the authors statethat they recognize that “many emotion researchersstudy only those emotions that can be conceptualizedas episodic, short-term responses to external stimuli”(this issue). It may have been more accurate had the au-thors said that “most” emotion theorists and research-ers view the experience of emotion as short-lived. This
widely adopted view is no arbitrary convention; it issupported by strong theoretical logic and empirical ev-idence (see Mandler, 1997, for a discussion). In short,baptizing the fledgling construct of caring in the murkyand troubled waters of contemporary emotion theoryand research is likely to be unhealthy for the constructand unpromising for our understanding of caregivingbehavior. We note in passing, however, that the au-thors’ definition of the “emotion” of caring—as “anenduring dyadic emotion that continues over the longterm, and that serves as an autonomous motivation tosee that the needs of a specific partner are met” (this is-sue)—sounds less like an emotion than it does an “atti-tude.” In any case, the venerable construct of attitudemay provide safer ground for the authors’ further de-velopment of the central construct of their model.
As all of the previously stated suggests, the Connec-tion Model is in need of further development. To saythat the source of caregiving behavior is the emotion ofcaring does not address the core questions any theoryof caregiving must answer and the predictions it mustmake: “Who will care for whom, and when and howand why?” To answer the motivational “why” questionwith “because they feel the emotion of caring” auto-matically prompts regress to the kind of question emo-tion researchers long have been struggling to answer:“Why do they feel the emotion [of caring; of fear; ofjoy]?” The antecedents of this emotion (the conse-quences having been specified by the authors to becaregiving behavior) must be specified. Is the predis-position to experience the emotion of caring innatelygiven? One foresees that arguments could be made thata caregiving behavioral syndrome may be an innate so-cial response system activated under certainspecifiable conditions. Or is caregiving learned? If so,what are the conditions conducive to such learning? Ifthe Connection Model can do no more than posit thatthe emotion of caring can be inferred from the fact ofcaregiving behavior, the model is dead on arrival. Itmay be noted that Baumeister and Leary (1995), whoposit a human “need to belong,” provided a useful tem-plate for the kind of evidence assembly and logicalanalysis that the introduction of a new motivationalconstruct requires.
In sum, the antecedents of the “emotion of caring”require specification and elaboration. Moreover, ifBell and Richard truly intend their Connection Modelto encompass caregiving behavior in all relationships,not simply the parent–infant relationship, there is agreat deal of relevant caregiving theory and researchbeyond the confines of the attachment literature to beconsidered and integrated. A Herculean task, to besure, but Bell and Richard’s goal is worth the effort. Ina nation in which concern is growing that many chil-dren are not receiving adequate care, a nation, further-more, in which the number of old and infirm isburgeoning, society’s expectation that psychology
108
COMMENTARIES
ought to be able to provide answers to a multitude ofcaregiving questions is likely to be voiced ever morefrequently and insistently in the years ahead.
Note
Ellen Berscheid, Department of Psychology, ElliottHall, University of Minnesota, Minneapolis, MN 55455.
References
Batson, C. D. (1998). Altruism and prosocial behavior. In D. T.Gilbert, S. T. Fiske, & G. Lindzey (Eds.),The handbook of so-cial psychology: Vol. 2.(4th ed., pp. 282–316). New York:McGraw-Hill.
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desirefor interpersonal attachment as a fundamental human motiva-tion. Psychological Bulletin, 117,497–529.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic Books.
Bowlby, J. (1988).A secure base. New York: Basic Books.Carnelley, K. B., Pietromonaco, P. R., & Jaffe, K. (1996). Attach-
ment, caregiving, and relationship functioning in couples: Ef-fects of self and partner.Personal Relationships, 3,257–277.
Cassidy, J., & Shaver, P. R. (Eds.). (1999).Handbook of attachment:Theory, research, and clinical applications.New York: Guilford.
Clark, M. S. (Ed.). (1991).Prosocial behavior: Review of person-ality and social psychology: Vol. 12.Newbury Park, CA:Sage.
Ekman, P., & Davidson, R. J. (Eds.). (1994).The nature of emotion.New York: Oxford University Press.
Feeney, J. A. (1996). Attachment, caregiving and marital satisfac-tion. Personal Relationships, 4,401–416.
Helgeson, V. S. (1994). Relation of agency and communion towell-being: Evidence and potential explanations.PsychologicalBulletin, 116,412–428.
Ickes, W. (Ed.). (1997).Empathic accuracy. New York: Guilford.Kunce, L. J., & Shaver, P. R. (1994). An attachment-theoretical ap-
proach to caregiving in romantic relationships. In K.Bartholomew & D. Perlman (Eds.),Advances in personal rela-tionships: Vol. 5. Attachment processes in adulthood(pp.205–237). London: Jessica Kingsley.
Mandler, G. (1997).Human nature explored. New York: OxfordUniversity Press.
Pierce, G. R., Lakey, B., Sarason, I., & Sarason, B. (1997).Sourcebookof social support and personality. New York: Plenum.
Rholes, W. S., Simpson, J. A., & Orina, M. M. (1999). Attachmentand anger in an anxiety-provoking situation.Journal of Person-ality and Social Psychology, 76,940–957.
Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Supportseeking and support giving within couples in an anxi-ety-provoking situation: The role of attachment styles.Journal of Personality and Social Psychology, 62,434–466.
Attachment Theory and Caregiving
Phillip R. Shaver and R. Chris FraleyDepartment of Psychology
University of California, Davis
Normally when one is asked to comment on a scien-tific article, the authors of the article have made a co-herent case of some kind, have rooted it in existingtheoretical and empirical literatures, and have sug-gested new directions for research. In the case ofBell and Richard’s “Forgotten Element” article, it isimpossible to adopt the usual approach, because thearticle contains so many mistakes, misunderstand-ings, and false leads that there is no coherent thesisto be discussed. Moreover, there are no concretesuggestions for research based on the authors’ theo-retical proposals. Our approach, therefore, will be tohighlight and correct some of the misconceptions inthe target article and evaluate the authors’ proposedsolutions to the problems they think they have iden-tified. Our comments will be organized according toBell and Richard’s four major criticisms of attach-ment theory’s conceptualization of caregiving.
Motivation for Caregiving
Bell and Richard erroneously assert that the motiva-tion for caregiving, according to attachment theory, is
“self-interest.” To our knowledge, caregiving hasnever been described by attachment theorists as moti-vated by self-interest, except in the sense that thecaregiving system, like all biobehavioral adaptations,evolved because it increased parents’ inclusive fitness.Attachment theory’s concepts ofsensitiveandrespon-siveparenting clearly point to a parent’s focus on his orher child’s needs and signals, not on the parent’s moredirectly selfish concerns. Every extended discussion ofattachment theory includes a list of some of the stimulithat initially elicit parental caregiving, such as the in-fant’s rounded features, large head and eyes, obviousvulnerability, crying when distressed, and other ex-pressions of emotion. In addition to exhibiting thesesimple qualities, infants evoke in parents more com-plex thoughts and feelings about such issues as thechild’s need for nourishment, for protection from dan-ger, for stimulating entertainment, for educational sup-port and guidance, and so on, which are often acted onwithout the infant having to produce any particularemergency signals. To the extent that Bell and Richardattempt to explain what causes “caring,” they seem tohave some of these same kinds of things in mind. Cer-
109
COMMENTARIES
tainly George and Solomon, the attachment theoristswho have written most extensively about thecaregiving system (e.g., George & Solomon, 1989,1996, 1999; Solomon & George, 1996, 1999), havetalked about parents’ pervasive concern, foresight,care, and responsibility and about the inevitable con-flicts between caregiving and other activities that ulti-mately relate to inclusive fitness. It is therefore aserious mistake for Bell and Richards to act as if at-tachment theorists who study parent–child relation-ships need a lecture on such matters.
With respect to using attachment theory in the studyof adult romantic, or pair-bond, relationships, Kunceand Shaver (1994) provided empirical evidence con-cerning differences between self-protection andcaregiving, clearly indicating that caregiving is not“selfish” in the everyday sense. (See also studies byCarnelley, Pietromonaco, & Jaffe, 1996, and Feeney,1996.) It is, instead, oriented toward a relationshippartner’s needs and wishes, including some that thepartner may not perceive as accurately as the caregiverdoes. Thus, there is no basis in either the literature onchild–parent attachment or the literature on romanticattachment for claiming that attachment theorists viewcaregiving as “selfish.”
In contrast to their false characterization of attach-ment theorists as attributing caregiving to self-inter-est, Bell and Richard propose that the emotion of car-ing plays a central role in motivating caregivingbehavior: “We assume that caregiving behavior arisesfrom an autonomous motivation that can successfullycompete with self-interest. We find that competingmotivation in the emotion of caring” (this issue). Thisexplanation strikes us as circular: Caregivers providecare because they care. Although Bell and Richardbeef up their description of caring (e.g., by highlight-ing the fact that genuine caring involves a focus onthe other’s needs rather than one’s own), nothing inthese additional comments allows them to escape thefundamental circularity of their conceptual frame-work. There is no explicit cause of care in their anal-ysis (except the infant’s responsiveness, to whichthey devote a small amount of attention); it is, as theysay, “autonomous.” In contrast, attachment theoryportrays care and other emotional experiences ofcaregiving as part and parcel of the caregiving sys-tem, a system involving biologically evolved goals,appraisals, actions, and affects.
Bell and Richard also say that, in contrast to meet-ing the parent’s own needs, the “set goal of the care-giver is to meet the needs of the dependent” (thisissue). This is not different from attachment theorists’proposals. Bowlby (1969/1982) believed that the func-tional outcome of caregiving was protection (a posi-tion elaborated by George & Solomon, 1999); howwould this be accomplished without the caregiver’s at-tending to and being sensitive to the infant’s needs and
vulnerability to various dangers? The theory’s empha-sis on the caregiver’s interest in meeting the needs ofwhat Bell and Richard call “the dependent” is the rea-son for directing so much research at the concepts ofsensitivityandresponsiveness.What would such termsrefer to if they did not imply being attuned to the needsof the infant and trying to meet those needs appropri-ately?
The Place of Emotion in Caregivingand Attachment
Bell and Richard believe that emotions are tightlywoven into the fabric of caregiving, and they criticizeBowlby for failing to incorporate emotions into histheoretical explanation of attachment and caregiving.It is important to note, however, that emotions occupya key role in all of Bowlby’s writings (notice the wordsanger, sadness,anddepressionin the titles of his sec-ond and third volumes, and the wordsecurityin the ti-tle of one of his later books; Bowlby, 1988). Bell andRichard are correct in saying that Bowlby did not givethese emotions “autonomy” or causal primacy in histheory. One reason is that emotions such as care, sepa-ration anxiety, and grief are important parts of what hewas trying to explain. As such, they could not them-selves be conceptualized as autonomous causes.
Bowlby (1969/1982) adopted ethologists’ conceptof behavioral systemspartly to clarify the nature of at-tachment-related emotional experiences. He believedthat emotions were signals or outputs from complexappraisal mechanisms and behavioral control mecha-nisms that evolved to serve survival- and reproduc-tion-related functions. According to this analysis, theattachment system is partly responsible for monitoringthe environment for cues concerning the availabilityand responsiveness of the caregiver. When the child’sgoals for proximity maintenance are violated, a suite ofbehaviors—and the physiological changes necessaryto execute those behaviors—is activated to restoreproximity. Some aspects of this “appraisal leading tobehavioral activation” process arefelt (to useBowlby’s term), and they are sometimes interpretedand labeled as emotions. But the feelings themselvesare part of a more complex process, not the initiators ofthat process.
The behavioral systems approach helps to explainemotional experiences by specifying the functions ofthe systems that generate them, the proximate and dis-tal goals of these systems, and the nature of the ap-praisal processes, action patterns, terminatingconditions, and behavioral feedback pathways thatgovern them. According to Bowlby (1969/1982, chap.7), the feelings that accompany the operation of behav-ioral systems are important in self-understanding andin communication with other people (including clini-
110
COMMENTARIES
cians). They are sometimes used as a kind of shorthandfor explaining how a person (including oneself) re-acted in a particular situation. But for the most part, thesubjective (felt) aspects of these reactions are not inthemselves viewed as causal.
Interestingly, Bowlby’s (1969/1982) analysis ofemotions foreshadowed contemporary emotion theo-ries (e.g., Ekman, 1992; Frijda, 1986; Lazarus, 1991;Oatley & Jenkins, 1996; Shaver, Schwartz, Kirson, &O’Connor, 1987), all of which view emotions as elic-ited by appraisals, which may or may not be conscious,and as comprising action tendencies supported by req-uisite physiological and attentional changes (Frijda,1986). The Consensus Model of emotions (Fischer,Shaver, & Carnochan, 1990) does not characterize thefelt aspects of emotions always as necessary or, inthemselves, causal. (In other words, our everyday pro-pensity to characterize emotions as “feelings” is some-what misleading. All scientific analyses are forced todeal with appraisals, physiological changes, andchanges in action tendencies and patterns of expres-sion and behavior that go well beyond “feelings,” andin fact are among the processes that are subjectively“felt.”) When emotions are viewed as changes in ac-tion tendencies (sometimes accompanied by feelings,sometimes not) resulting from appraisals of events inrelation to goals and concerns (Frijda, 1986; Oatley &Jenkins, 1996), one can see why proximity seeking,cuddling, providing care, and so on are accompaniedby strong feelings. One can also see why the feelingsthemselves are not given an autonomous or causallyprior role in attachment theory. These subjectively feltemotions are aspects of the operation of behavioralsystems, not the autonomous triggers of those systems.In other words, to return specifically to caregiving, car-ing does not explain the caregiving behavioral system;the caregiving system explains the feelings Bell andRichard labelcaring.
How do Bell and Richard go about putting the, tothem, forgotten emotion of care back into caregiving?Somewhat carelessly. In their words, “cognitions [bywhich we assume they mean what Lazarus and othercontemporary emotion theorists call ‘appraisals,’ aterm deliberately connotingevaluation, not simplycool perception] are not what motivate the parent’s ac-tions—emotions are, and more basically the emotionof caring” (this issue). This proposed solution to whatwe believe is a nonexistent problem—once one under-stands both attachment theory and contemporary theo-ries of emotion—raises problems. First, part of whatwe wish to explain, namely the feelings and behaviorsof caring, becomes a proposed cause of itself: Peoplecare because they care. Second, this phenomenologicalghost is inserted into the caregiving machine withoutany reference to an explicit theory of emotion. Bell andRichard do not mention modern appraisal theories ofemotion, nor do they rely on some alternative theory.
Instead, they seem to rely on their own strong sensethat because care feels so moving and important, itmust in fact be a causal prime mover, an autonomousforce.
Attachment and Caregiving “Bonds”
Bell and Richard’s section on affectional bonds isone of the most confusing sections in their article. Theauthors appear to argue that there are too many ambi-guities in the concepts ofaffectional bondandattach-ment bond,including the purported implication thatbonds develop exceedingly quickly, the connection be-tween affection and attachment, and the degree towhich people “choose” to become bonded. In our opin-ion, Bell and Richard have read too much into the“bond” construct. Like attachment, the termaffectional bondsis metaphorical (there being no lit-eral, physical “bond”). It refers to the fact that two indi-viduals are “tied” or “bound” to each other in arelationship. When two people are bound into an at-tachment relationship, they recognize each other asuniquely important; monitor each other’s where-abouts; have some understanding of each other’s goals,intentions, and desires (with sufficient age or cognitivedevelopment); have experienced strong emotions inrelation to each other; and would be extremely upsetand, at least for a time, lost without each other. Thiscomplex situation is summarized, for convenience, bythe termbond.
How do such bonds form? Contrary to what Belland Richard believe, attachment theorists do not arguethat affectional or attachment bonds form quickly, in away similar to imprinting. Indeed, Bowlby(1969/1982) was cautious about using the termim-printing, a term originally designed to describe the at-tachment behavior of birds, when describingattachment processes in humans. Because attachmentbehavior appears to have evolved independently inbirds and mammals, Bowlby acknowledged the possi-bility that the neural machinery underlying attachmentbehavior in the two taxa may be quite different. Al-though the process of attachment begins early in devel-opment (Bowlby, 1969/1982), many researchers (e.g.,Hazan & Zeifman, 1994; Marvin & Britner, 1999;Scheper-Hughes, 1992) described the gradual processby which an infant becomes attached to a particularparent and how the caregiver becomes increasingly in-vested in a particular offspring. Similarly, research onattachment development in romantic relationships sug-gests that the process of attachment formation oftentakes as long as 2 years (Fraley & Davis, 1997; Hazan& Zeifman, 1994).
In both kinds of relational contexts, the process ofbond formation involves conscious and unconsciousfamiliarity; numerous experiences of synchronous, in-
111
COMMENTARIES
terdependent interactions; intertwined goals and pro-jects; and perhaps (on the part of adults) cognitivedissonance reduction. In some ways, the process issimilar from both sides of a child–parent relationship,but because the concerns, goals, representations, andso on are far from identical on the two sides, some ex-periences, such as loss, probably differ for “depend-ents” and “caregivers.” The dependents are likely tofeel anxious and unprotected; the caregivers are likelyto experience guilt and a continuing wish to be protec-tive and helpful (Bowlby, 1980). Each position re-quires a massive reorganization of personal goals, butthe nature of the goals is not the same.
How do Bell and Richard revise current views ofcaregiving and attachment bonds? First, they suggestthat caregiving needs to be conceptualized as moreproactive than reactive: “In contrast [to attachmenttheory], connection theory [i.e., the authors’ own the-ory] views caregiving as more proactive than reactive”(this issue). There is nothing, however, about contem-porary perspectives on caregiving to imply thatcaregiving is simply a reactive process. As George andSolomon (1999) explained, using both human and non-human primate examples, the caregiving system is ac-tive much of the time, even when the dependent showsno signs of realizing that he or she is in danger or inneed. Parental interventions are not viewed as always,or even usually, being responses to explicit signals on adependent’s part. Moreover, attachment researchers(e.g., Slade et al., 1995) agreed with Bell and Richardthat caregiving working models often develop before achild is born (e.g., during pregnancy). In fact, Sladeand her colleagues already showed that a woman’sown attachment working models influence the natureof her imagined relationship with the child-to-be.
Bell and Richard also argue that we need a betterexplanation of the enduring nature of affectionalbonds: “Most attachment researchers describe the at-tachment and caregiving bonds as long-lasting and ir-replaceable, but this is a characteristic of affectivebonds that is described (and assumed) rather than ex-plained” (this issue). Attachment theory views thebonds as serving the survival and reproductive goals(i.e., the goals of inclusive fitness) of both parents andchildren. The bonds are long-lasting on the part of par-ents because parental investment and parental effortcan last a lifetime. They are long-lasting on the part ofchildren because parents continue to serve as safe ha-vens and secure bases for many years, sometimes for aslong as the parents live. Of course, saying this is not thesame as explaining what it is about the underlying neu-ral networks that keep an “affectional bond” in place,sometimes long after a partner has died, and this is aninteresting question for research.
What is Bell and Richard’s proposed explanatorymechanism? They say, “The caregiving is the bond.…The caregiving bond endures because it is the enduring
emotion of caring” (this issue). Not only does this ex-planation trivialize the metaphor of affectional bonds,it is circular: The bond is enduring because the bond iscaring and caring is enduring.
Explaining Sensitivity andResponsiveness
Bell and Richard criticize attachment theory forfailing to provide an explanation of why responsiveand sensitive caregivers behave as they do. Attachmentresearchers, however, have expended considerable en-ergy exploring the transmission of caregiving behav-iors from parents to children (see, e.g., De Wolff & vanIJzendoorn, 1997; George & Solomon, 1999; vanIJzendoorn, 1995). Many researchers believe that theresponsiveness of the caregiver to the child’s needshelps the child learn to regulate his or her emotions,and that the patterns of interaction (i.e., the contingen-cies, responses, and “rules” of interaction) between theparent and child become assimilated into a child’s de-veloping representations of the world. These “workingmodels” are carried forward into young adulthood and,in conjunction with other factors correlated with work-ing models, such as marital support (Belsky,Rosenberger, & Crnic, 1995; Crowell, Fraley, &Shaver, 1999), come to shape the way a person feelsabout his or her own child before it is born (Fonagy,Steele, & Steele, 1991; Slade et al., 1995). Further,these working models manifest themselves in the waysin which new parents organize their caregiving behav-ior toward their own children. As George and Solomon(1999) noted, parents may sometimes be preoccupiedwith their own fears or insecurities about parenting,and this interferes with their ability to sensitively orflexibly meet the needs of their children. Bell andRichard believe that this explanation constitutes a“moral exhortation” (this issue) and “oversimplifiesand further trivializes the process of caregiving” (thisissue), but we view it as a rather sophisticated, if in-complete, explanation of variation in caregiving be-havior.
The authors’ alternative is to explain the caregiver’smotivation for caregiving via empathy and responsi-bility: “Because caring represents a concern for the de-pendent’s needs, empathy follows from caring and is amechanism through which the caring is implemented”(this issue). This explanation strikes us as vacuous. Itdoes not tell us why some caregivers are less investedin their offspring than others are (although it does rec-ognize that caregiving can be exhibited in variable de-grees), how exactly this variation may be interpreted orexperienced by the child, and how regularities incaregiving patterns may affect the subsequent person-ality development of both relationship partners. Fur-ther, it seems peculiar to us that the concepts of
112
COMMENTARIES
empathyand responsibilitybear such striking resem-blance to the concepts ofsensitivityand responsive-nessoffered by the traditional attachment perspective.In this respect, it is difficult to separate Bell and Rich-ard’s supposed contribution from existing theory.
Concluding Comments
Bell and Richard’s critique of attachment theorists’conception of caregiving seems uninformed and mis-guided. Their alternative view offers little of value.Theirapproachseemstobemotivatedprimarilybyade-sire to incorporate the potency of feelings into prevail-ing views of caregiving. But there is nothing in contem-porary theories to dissuade readers from noticing,celebrating, or appreciating the subjective side ofcaregiving. Bell and Richard seem to believe that itmakes more sense to derive empathy and responsibilityfrom “care” than it does to incorporate sensitivity-em-pathy and responsiveness-responsibility into the nor-mal workings of the caregiving behavioral system. Forthem, the absence in attachment theory of an emphasison the phenomenology of care, which they attempt tocapture in their “I feel wonderful” paragraph (this is-sue), means that attachment theory is leaving out some-thing important. But readers of George and Solomon’s(1999) or Kunce and Shaver’s (1994) attachment theo-reticalperspectivesoncaregivingwill notice that the ro-mance, the feeling, is all there, and in a much more com-plete and sophisticated theoretical context than the oneprovided by Bell and Richard. George and Solomon’sparent–child and Kunce and Shaver’s romantic-rela-tionship conceptions of caregiving have already led tothe creation of useful measures and to both theoreticallyand clinically significant empirical discoveries. In con-trast, Bell and Richard offer no proposals for researchbased on their ideas. In this way, their Connection The-ory definitely differs from attachment theory.
Note
Phillip R. Shaver and R. Chris Fraley, Departmentof Psychology, University of California, Davis,CA 95616–8686. E-mail: [email protected]
References
Belsky, J., Rosenberger, K., & Crnic, K. (1995). The origins of at-tachment security: “Classical” and contextual determinants. InS. Goldberg, R. Muir, & J. Kerr (Eds.),Attachment theory: So-cial, developmental, and clinical perspectives(pp. 153–183).Hillsdale, NJ: Analytic Press.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic Books.
Bowlby, J. (1980).Attachment and loss: Vol. 3. Loss: Sadness anddepression. New York: Basic Books.
Bowlby, J. (1988).A secure base. New York: Basic Books.Carnelley, K. B., Pietromonaco, P. R., & Jaffe, K. (1996). Attach-
ment, caregiving, and relationship functioning in couples:Effects of self and partner.Personal Relationships, 3,257–277.
Crowell, J., Fraley, R. C., & Shaver, P. R. (1999). Measures of indi-vidual differences in adolescent and adult attachment. In J.Cassidy & P. R. Shaver (Eds.),Handbook of attachment: The-ory, research, and clinical applications(pp. 434–465). NewYork: Guilford.
De Wolff, M. S., & van IJzendoorn, M. H. (1997). Sensitivity and at-tachment: A meta-analysis on parental antecedents of infant at-tachment.Child Development, 68,571–591.
Ekman, P. (1992). An argument for basic emotions.Cognition andEmotion, 6,169–200.
Feeney, J. A. (1996). Attachment, caregiving, and marital satisfac-tion. Personal Relationships, 3,401–416.
Fischer, K. W., Shaver, P. R., & Carnochan, P. (1990). How emotionsdevelop and how they organize development.Cognition andEmotion, 4,81–127.
Fonagy, P., Steele, H., & Steele, M. (1991). Maternal representationsof attachment during pregnancy predict organization of in-fant–mother attachment at one year of age. Child Development,62,891–905.
Fraley, R. C., & Davis, K. E. (1997). Attachment formation and trans-fer in young adults’ close friendships and romantic relation-ships.Personal Relationships, 4,131–144.
Frijda, N. H. (1986).The emotions. Cambridge, England: CambridgeUniversity Press.
George, C., & Solomon, J. (1989). Internal working models ofcaregiving and security of attachment at age six.Infant MentalHealth Journal, 10,222–237.
George, C., & Solomon, J. (1996). Representational models of rela-tionships: Links between caregiving and attachment.InfantMental Health Journal, 17,198–216.
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. 649–670). New York:Guilford.
Hazan, C., & Zeifman, D. (1994). Sex and the psychological tether. InK. Bartholomew & D. Perlman (Eds.),Advances in personal re-lationships: Vol. 5. Attachment process in adulthood(pp.151–178). London: Jessica Kingsley.
Kunce, L. J., & Shaver, P. R. (1994). An attachment-theoretical ap-proach to caregiving in romantic relationships. In K.Bartholomew & D. Perlman (Eds.),Advances in personal rela-tionships: Vol. 5. Attachment processes in adulthood(pp.205–237). London: Jessica Kingsley.
Lazarus, R. S. (1991).Emotion and adaptation. New York: OxfordUniversity Press.
Marvin, R. S., & Britner, P. A. (1999). Normative development: Theontogeny of attachment. In J. Cassidy & P. R. Shaver (Eds.),Handbook of attachment: Theory, research, and clinical appli-cations(pp. 44–67). New York: Guilford.
Oatley, K., & Jenkins, J. M. (1996).Understanding emotions. Cam-bridge, MA: Blackwell.
Scheper-Hughes, N. (1992).Death without weeping: The violence ofeveryday life in Brazil. Berkeley: University of California Press.
Shaver, P. R., Schwartz, J., Kirson, D., & O’Connor, C. (1987).Emotion knowledge: Further exploration of a prototype ap-proach.Journal of Personality and Social Psychology, 52,1061–1086.
Slade, A., Dermer, M., Gerber, J., Gibson, L., Gaf, F., Siegel, N., &Tobias, K. (1995, March).Prenatal representation, dyadic in-teraction, and quality of attachment. Paper presented at the bi-
113
COMMENTARIES
ennial meeting of the Society for Research in Child Develop-ment, Indianapolis, IN.
Solomon, J., & George, C. (1996). Defining the caregiving system:Toward a theory of caregiving.Infant Mental Health Journal,17,183–197.
Solomon, J., & George, C. (1999). The caregiving behavioral sys-tem in mothers of infants: A comparison of divorcing and mar-
ried mothers. Attachment and Human Development, 1,171–190.
van IJzendoorn, M. H. (1995). Adult attachment representa-tions, parental responsiveness, and infant attachment: Ameta-analysis on the predictive validity of the Adult At-tachment Interview. Psychological Bulletin, 117,387–403.
Caregiving, Attachment Theory, and the Connection TheoreticalOrientation
Jeffry A. Simpson and W. Steven RholesDepartment of Psychology
Texas A&M University
Bell and Richard raise some important questionsabout how caregiving is—and theoretically shouldbe—related to basic attachment principles and pro-cesses. Until recently (see Belsky, Steinberg, &Draper, 1991; Chisholm, 1996; George & Solomon,1996, 1999), attachment theorists have not sufficientlyaddressed how the caregiving behavioral system inter-faces with the attachment system. In this commentary,we first discuss some reasons why caregiving appearsto be the “forgotten element” in attachment theory. Wethen discuss several misconceptions the authors haveabout attachment theory and indicate how recentontogenetic theories have attempted to link early pat-terns of attachment to subsequent mating strategiesand parenting styles over the life span. We conclude bydiscussing why the authors’ new theory is problematic,and we suggest that a different set of questions aboutthe theoretical connections between attachment andcaregiving needs to be asked.
Why Is Caregiving the “ForgottenElement?”
One of the reasons why attachment theory has notbeen well integrated with caregiving is that infantsurvival and successful childrearing are two distinctlife tasks that occur at different points in develop-ment and, according to evolutionary principles, canhave independent effects on reproductive fitness(Simpson, 1999). This is why most theorists concep-tualize attachment and caregiving as separate,evolved systems. Nonetheless, meager theoretical ad-vances also have hindered integration of the two sys-tems, especially since attachment theory and researchhas moved into the realm of adult romantic relation-ships. Bell and Richard are correct in noting that at-tachment theory focuses on two major goalstates—the need for proximity (in infants) and the
need for felt security (in older children and adults).These goal states have different implications for howone views the nature, purpose, and functions of at-tachment styles in infants versus adults. To com-pound matters, theorists have not clarified how thefunctions and objectives of caregiving directed toyoung, vulnerable infants are similar to (or differentfrom) more reciprocal forms of caregiving that occurbetween equal-status adults in mating relationships.As we discuss below, the nature, goals, and functionsof caregiving may be very different in different typesof relationships (e.g., parent–child, romantic, casualfriendships), which leads one to question whether theauthors’ new theory of caregiving operates in a simi-lar manner across different types of relationships.
Attachment scholars have also done a relativelypoor job of specifying how different working modelsare interconnected and should be associated withcaregiving (e.g., models of the parent–child relation-ship, models of romantic partners, models of closefriends; see Collins & Read, 1994). This lack of speci-ficity has confused and perhaps camouflaged some ofthe unique goals and functions that caregiving mayserve in different types of relationships. For example,the Adult Attachment Interview (AAI), which was de-veloped to assess patterns of attachment to one’s par-ents as assessed in adulthood (Hesse, 1999; Main,Kaplan, & Cassidy, 1985), was actually designed topredict the attachment styles of respondents’ infants inthe Strange Situation. As a result, the AAI containsvariance reflecting both memories and interpretationsof how individuals were cared for by their own parentsand how they care for their own children. The StrangeSituation, on the other hand, was developed to classifyattachment patterns in vulnerable, highly dependentinfants based on how they use their caregivers to regu-late their negative emotions when they are distressed.And most self-report measures of attachment to ro-mantic partners inquire about the seeking and giving of
114
COMMENTARIES
care (Shaver, Belsky, & Brennan, 2000), but in thecontext of reciprocal, equal-status sexual relationshipswhere partners jointly give and receive support. If thenature and functions of caregiving are somewhatunique across different types of relationships, one muststudy how multiple working models affect caregivingin different types of relationships to fully understandhow attachment relates to caregiving. With few excep-tions, attachment researchers have not done this.
Misconceptions About AttachmentTheory
Bell and Richard also paint an overly simplistic andsometimes misleading portrait of attachment theory.They claim, for example, that attachment theory viewsthe caregiving system as “purely reactive” to environ-mental triggers rather than being proactive in nature.Internal working models can and often do producespontaneous, proactive caregiving in the absence ofenvironmental activators, some of which are adaptive(e.g., when caregivers with secure working models an-ticipate and prevent their children from being harmed)and some of which are maladaptive (e.g., when care-givers with preoccupied models are chronically over-protective; see George & Solomon, 1999).
The authors also claim that attachment theory doesnot offer a “plausible motivation” for caregiving. Themotivation for caregiving, however, is explained byother middle-level evolutionary theories that focusmore directly on parental investment and caregiving,such as parent–offspring conflict theory (Trivers,1974), life history theory (Williams, 1966), and sexualselection and parental investment theory (Trivers,1972). Each of these theories is based on general as-sumptions from inclusive fitness theory (Hamilton,1964), which clearly explains why most parents shouldhave strong “self-interest” in investing heavily in theirchildren. It is important to recognize that attachmenttheory was originally developed to address infant sur-vival, and most of Bowlby’s major ideas were con-ceived well before other middle-level evolutionarytheories were developed (see Simpson, 1999). Recenttheoretical extensions involving felt security and at-tachment processes across the life span have begun todeal with issues that may be relevant to different lifetasks (i.e., mating and parenting). Considerable debatecurrently exists about what evolutionary functionsadult attachment styles may serve (see Kirkpatrick,1998; Zeifman & Hazan, 1997, for contrasting views).
Theauthorsalsocontendthatattachment theorydoesnot allow for affect to be a “causal force” in the affectregulation system. However, responses to either exter-nal threats or chronic worries induced by working mod-els can generate emotional states within individuals thatare capable of activating attachment systems and influ-
encing affect regulation processes (see Simpson &Rholes, 1994). Moreover, the factors acting as causesand effects within the attachment system become diffi-cult to differentiate once the system is activated.
Bell and Richard also suggest that no attach-ment-based theoretical work has tried to explain why“quality” caregivers behave as they do. AlthoughBowlby did not provide a direct explanation, severalrecent life history models that are grounded in attach-ment theory have done so (see the life-span models ofsocial development proposed by Belsky et al., 1991;Chisholm, 1996). These models integrate attachmenttheory, sexual selection theory, and life history theory.
Finally, the authors either misinterpret or oversim-plify several additional points. For instance, attach-ment theory does not suggest that no caring precedescaregiving; attachment bonds do not necessarily de-velop toward the “best” available caregiver; the attach-ment system is not simply switched on and off in anall-or-none fashion; the theory does not propose thatcaregivers respond “mechanically” to their children’scrying; and it does not assume that poor-quality par-ents are incompetent but not unmotivated. Most if notall of these assumptions are wrong, and they are not en-dorsed by most contemporary attachment theorists.
The Utility of the ConnectionTheoretical Orientation Model
Not all relationships are attachment relationships,and, even among those that are, there are many impor-tant facets of close relationships that attachment theorysimply does not address. It is important to remain fo-cused on the theoretical core of attachment theory thatBowlby, Ainsworth, Main, and others have estab-lished, if for no other reason than to ensure that attach-ment theory avoids becoming imprecise,overextended, and no longer capable of providing use-ful guidance to hypothesis generation and explanatorythinking (see Main, 1999).
Attachment relationships differ from other types ofrelationships in that attachment figures are persons towhom one turns in times of distress, mainly becausesuch persons are willing and able to promote one’ssafety or “felt security” (Main, 1999). However, thereoften is an inherent asymmetry in attachment relation-ships whereby one person has strong needs and viewshis or her partner as offering a possible “solution” to acurrent problem. In adult–child relationships, thisasymmetry is witnessed in the fact that the adult oftenserves as the child’s solution, but not vice versa. Inadult–adult relationships, temporary periods of asym-metry occur when one partner views the other as“stronger and wiser” (Shaver, Hazan, & Bradshaw,1988) and as the likely solution to his or her immediateconcerns or distress.
115
COMMENTARIES
In light of these considerations, how shouldcaregiving be defined from the vantage point of attach-ment theory? Caregiving entails the willingness to ac-cept (rather than turn away) the dependency of othersand to respond in ways that either reduce their distressor enhance their safety. Given this definition, there aremany forms and modes of caregiving that simply arenot relevant to attachment theory. Because it is essen-tial to concentrate on the most central, distinguishingfeatures of an attachment perspective on caregiving,the connection theoretical orientation proposed by Belland Richard is problematic primarily because it is tooinclusive.
Are the Right Questions Being Asked?
Bell and Richard state that the key questions aboutcaregiving for attachment theory should center on: (a)identifying the motivation for caregiving, (b) finding aplace for emotion within caregiving, (c) describing thenature of attachment and caregiving bonds, and (d)providing an explanation for caregivers’ responsive-ness and sensitivity. We do not dispute the utility orimportance of these questions, but we believe that themost critical issues are not addressed in Bell and Rich-ard’s article.
At present, there is ample evidence that adults’ rep-resentations of their childhood experiences with theirparents affect the patterns of attachment that their chil-dren develop with them. A portion of this effect is me-diated through maternal sensitivity and responsiveness(van IJzendoorn, 1995). However, the routes by whichchildhood experiences become translated into differentpatterns of adult caregiving (and, in turn, different in-fant attachment patterns) remain largely unknown. Afull explanation of this process will require the identifi-cation of behaviors that distinguish persons with dif-ferent types of early attachment experiences.Following this, one must establish links between thesebehaviors and infant attachment patterns, after whichone must explain how mental representations of earlyexperiences influence adult caregiving practices. Co-gent answers to these enigmas are, in our view, themost interesting and the most pressing. Unlike mosttheories of early experience, attachment theory doesnot argue that one’s experiences in early relationshipsnecessarily cause subsequent behavior; instead, one’srepresentation of experiences (apparent in memoriesand attributions) is the causal element (see Main,1999). According to this perspective, one may have ad-verse attachment experiences in childhood yet revealno impact of such experiences in one’s caregiving orparenting if one’s representations of these experiencesare “healthy” (e.g., problems are acknowledged, butparents were forgiven). Consistent with this viewpoint,Cohn, Cowan, Cowan, and Pearson (1992) found thatparents who display “earned” security (i.e., those with
difficult childhood experiences but who developedhealthy, secure relationships with their parents) do notdiffer from persons with healthy, largely positivechildhood experiences in terms of how they care fortheir own infant children. What is needed is a theoreti-cal framework that can explain the processes throughwhich internal working models affect parents’caregiving and their infants’ responses to them.
In summary, we agree that attachment theory’s ac-count of caregiving is not well developed. However,we question whether the connection theoretical orien-tation will appreciably advance our understanding ofeither attachment theory or caregiving.
Note
Jeffry A. Simpson, Department of Psychology,Texas A&M University, College Station, TX77843–4235. E-mail: [email protected]
References
Belsky, J., Steinberg, L., & Draper, P. (1991). Childhood experience,interpersonal development, and reproductive strategy: An evolu-tionary theory of socialization.Child Development, 62,647–670.
Chisholm, J. S. (1996). The evolutionary ecology of attachment orga-nization.Human Nature, 7,1–38.
Cohn, D., Cowan, P., Cowan, P., & Pearson, J. (1992). Mothers’ andfathers’ working models of childhood attachment relationships,parenting style, and child behavior.Development andPsychopathology, 4,417–431.
Collins, N. L., & Read, S. J. (1994). Cognitive representations of at-tachment: The structure and function of working models. In K.Bartholomew & D. Perlman (Eds.),Attachment processes inadulthood(pp. 53–90). London: Jessica Kingsley.
George, C., & Solomon, J. (1996). Representational models of rela-tionships: Links between caregiving and attachment.InfantMental Health Journal, 17,198–216.
George, C., & Solomon, J. (1999). Attachment and caregiving. In J.Cassidy & P. R. Shaver (Eds.),Handbook of attachment: The-ory, research, and clinical applications(pp. 649–670). NewYork: Guilford.
Hamilton, W. D. (1964). The genetical evolution of social behaviour.Journal of Theoretical Biology, 7,1–52.
Hesse, E. (1999). The Adult Attachment Interview: Historical andcurrent perspectives. In J. Cassidy & P. R. Shaver (Eds.),Hand-book of attachment: Theory, research, and clinical applications(pp. 395–433). New York: Guilford.
Kirkpatrick, L. A. (1998). Evolution, pair-bonding, and reproduc-tive strategies: A reconceptualization of adult attachment. InJ. A. Simpson & W. S. Rholes (Eds.),Attachment theory andclose relationships(pp. 353–393). New York: Guilford.
Main, M. (1999). Attachment theory: Eighteen points with sugges-tions for future research. In J. Cassidy & P. R. Shaver (Eds.),Handbook of attachment: Theory, research, and clinical appli-cations(pp. 845–887). New York: Guilford.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, child-hood, and adulthood: A move to the level of representation. In I.Bretherton & E. Waters (Eds.), Growing points of attachment the-ory and research (pp. 66–104).Monographs of the Society for Re-search in Child Development, 50(1-2, Serial No. 209).
Shaver, P. R., Belsky, J., & Brennan, K. A. (2000). The Adult AttachmentInterview and self-reports of romantic attachment: Associationsacross domains and methods.Personal Relationships, 7,25–43.
116
COMMENTARIES
Shaver, P. R., Hazan, C., & Bradshaw, D. (1988). Love as attach-ment: The integration of three behavioral systems. In R. J. Stern-berg & M. L. Barnes (Eds.),The psychology of love(pp. 68–99).New Haven, CT: Yale University Press.
Simpson, J. A. (1999). Attachment theory in modern evolutionaryperspective. In J. Cassidy & P. R. Shaver (Eds.),Handbook ofattachment: Theory, research, and clinical applications(pp.115–140). New York: Guilford.
Simpson, J. A., & Rholes, W. S. (1994). Stress and secure base rela-tionships in adulthood. In K. Bartholomew & D. Perlman (Eds.),Attachment processes in adulthood(pp. 181–204). London:Jessica Kingsley.
Trivers, R. L. (1972). Parental investment and sexual selec-tion. In B. Campbell (Ed.),Sexual selection and the de-
scent of man, 1871–1971(pp. 136–179). Chicago: Al-dine-Atherton.
Trivers, R. L. (1974). Parent–offspring conflict.American Zoologist,14,249–264.
van IJzendoorn, M. (1995). Adult attachment representations, paren-tal responsiveness, and infant attachment: A meta-analysis onthe predictive validity of the Adult Attachment Interview.Psy-chological Bulletin, 117,387–403.
Williams, G. (1966).Adaptation and natural selection. Princeton,NJ: Princeton University Press.
Zeifman, D., & Hazan, C. (1997). Attachment: The bond inpair-bonds. In J. A. Simpson & D. T. Kenrick (Eds.),Evolution-ary social psychology(pp. 237–263). Mahwah, NJ: LawrenceErlbaum Associates, Inc.
Attachment, Intimacy, and Generativity
Dan P. McAdamsProgram in Human Development and Social Policy
Northwestern University
As Bell and Richard point out in the last para-graph of their very fine article on caregiving and at-tachment, Bowlby was quite fond of militarymetaphors. In its cybernetic intricacy, the attach-ment system was akin, Bowlby (1969) suggested, toan “antiaircraft gun.” Constantly monitoring feed-back in relation to a set goal or target, the attachmentsystem aimed at shooting down the many predatorsand dangers that threatened the helpless infant in thePleistocene environment of evolutionaryadaptedness. A metaphor that Bowlby liked evenbetter wasthe secure base. In his last book,A SecureBase, Bowlby (1988) wrote that the attachment bondideally provides a base
from which a child or an adolescent can makesortiesinto the outside world and to which he can return know-ing for sure that he will be welcomed when he gets there,nourished physically and emotionally, comforted if dis-tressed, reassured iffrightened. In essence this role isone ofbeing available,ready to respond when called onto encourage and perhaps assist, but to intervene ac-tively only when clearly necessary. In these respects it isa role similar to that of theofficer commanding a mili-tary basefrom which anexpeditionary forcesets outand to which it canretreat should it meet asetback.Much of the time the role of the base is a waiting one butit is nonetheless vital for that. For it is only when the of-ficer commanding the expeditionary force is confidenthis base is secure that he dares press forward and takerisks. (p. 11, [italics added throughout])
Bell and Richard should be commended for soft-ening attachment’s war rhetoric with the discourse ofcaring. More important, their carefully reasoned and
insightful exegesis of the attachment bond reveals thesingularity of this primal relationship. Armed with afashionable theory and with quick-and-easy pa-per-and-pencil measures to assess individual differ-ences in attachment security, social psychologistshave rushed ahead to implicate attachment as theguiding process for a number of different kinds of in-terpersonal relationships, including adult friendshipsand romantic love. It became virtually accepted wis-dom among many social psychologists and cliniciansthat attachment processes are central in the develop-ment of intimate relationships between adults (Reis &Patrick, 1996). Bell and Richard, however, urge cau-tion. Before we extend the attachment metaphor tothe four corners of the interpersonal universe, weneed to take a closer look at the complex dynamics ofthis unique bond. To elaborate on two important dis-tinctions that come out of Bell and Richard’s analy-sis, let me borrow terminology from Erik Erikson’s(1963) well-known stage model of psychosocial de-velopment. Bell and Richard’s account suggests thatthe attachment bond is not the same thing as adult in-timacy (Erikson’s Stage 6) and that the caregivingcomponent of attachment is not the same thing asgenerativity (Erikson’s Stage 7). In our quest to con-nect concepts with different pedigrees and connota-tions, we need to slow down, Bell and Richard tell us,lest we blur important theoretical distinctions.
Attachment Versus Intimacy
Bowlby’s (1988) military metaphors are revealing.In the prototypical scenario of caregiver–infant attach-ment, a weaker partner (the infant, or what Bell and
117
COMMENTARIES
Richard term thedependent) faces a threatening world(predators, strangers—the enemy). The weaker partnerreceives sustenance and protection from the strongerpartner (theattachmentobject, orwhatBell andRichardterm thecaregiver), who provides a haven of safety anda secure base from which exploratory “sorties” can beconducted. The stronger partner functions as a “com-manding officer” of sorts, holding considerably morepower in this relationship. Nonetheless, she or he is as-tute enough to wait it out most of the time and to simply“be available” should the weaker partner meet a “set-back.” In an impressive theoretical advance, Bell andRichard argue for a more activist caregiver—a com-manding officer who, motivated by the enduring dyadicemotionofcaring,adoptsa rathermoreproactivestancevis-à-vis the infant, seeking first and foremost to meetthedependent’sneedsand tocontinue tomeet themwellinto the future. Nonetheless, Bell and Richard agreewith Bowlby in characterizing the attachment relation-ship as one of relative unequals. There is a “hierarchybetween parent and child,” they write. “The parent con-trols virtually all of the resources in the parent–child re-lationship” (this issue). The parent is bigger, stronger,older, wiser, richer, and more self-aware. According toBell and Richard, furthermore, the parent is gifted withempathy and responsibility, which serve as psychologi-cal mechanisms through which caring is implemented.The caregiver feels the desire to know the dependent(empathy) and to meet the dependent’s needs (responsi-bility).Thecaregiver (bydefinition)cares for the infant.When it comes to caring, the infant, by contrast, is clue-less.
Unless we wish to argue that one member of the ma-ture adult friendship or romantic relationship is peren-nially the clueless dependent, we would do well to holdback our enthusiasm about attachment as a model forromantic love and intimate relations. Bell and Richardmaintain that attachment is different from romanticlove. It is also different from intimacy. Intimacy is aquality of interpersonal relating through which part-ners share personal thoughts, feelings, and other im-portant aspects of themselves with each other. In anexhaustive review of the construct, Prager (1995)wrote, “all conceptions of intimate interactions seem tocenter on the notion that intimate behavior consists ofsharing that which is personal” (pp. 20–21). McAdams(1989) sees “the sharing of one’s innermost self” as thecardinal feature of intimacy (p. 49). McAdams con-tends that an idealized model for intimate exchange iswhat Martin Buber (1970) called the “I–Thou encoun-ter.” The I–Thou is a special quality of interpersonalexperience in which the I (self) and Thou (other) sharewith each other personal thoughts, feelings, observa-tions, and so forth in such an intensive manner thateach becomes visible to the other in his or her whole-ness and unique individuality. What mainly distin-guishes intimacy from attachment, then, is intimacy’s
fierce insistence on an intensive egalitarian exchangebetween two autonomous selves. In emotional terms,furthermore, whereas attachment assuages fear and of-fers protection, intimacy assuages loneliness and of-fers the opportunity to further extend one’sunderstanding of self and other (Sullivan, 1953). Inwhat Stern (1985) characterizes as nonverbal statesharing, infants and their caregivers may engage in arudimentary form of intimacy. But attachment pro-cesses would appear to dominate the psychosocialscene in the first few years of life, whereas intimacyawaits a more fully developed sense of self and theachievements in cognitive development and role tak-ing that attend later childhood and adolescence (Hart,1988; Selman, 1980; Sullivan, 1953).
Intimacy is an important aspect of the phenomenol-ogyof romantic love inadulthood.SternbergandGrajek(1984) identify intimacy, passion, and commitment ascentral components of romantic love. Attachment pro-cesses surely play a part, too (Hazan & Shaver, 1987).But in many mature relationships between partners whoperceive themselves as equals, the model of care-giver–dependent is not an easy fit. In the adult attach-ment literature, Bell and Richard suggest, it has been“tacitly assumed that caregiving would occur as somekind of implicit exchange—each partner givingnurturance because each wanted to receive the other’snurturance” (this issue). Bell and Richard are probablyright, but I share their ambivalence about the adequacyof this kind of conceptualization in understanding whatgoes on between lovers. Although caregiving is at thecenter of attachment, it would appear to be only one ofmany different aspects of mature romantic love, and insome couples a minor aspect at that.
Attachment Versus Generativity
Caregiving begins with caring, Bell and Richardpoint out. Although this contention may seem trivial, itproves to be theoretically very important. Caring is “anenduring dyadic emotion that continues over the longterm and that serves as an autonomous motivation to seethat the needs of a specific partner are met” (this issue).Once the caregiver imagines or perceives that the infantis (or will be) responsive to him or her, the caregiver be-gins to feel theemotionofcaring.Fromthestandpointofthe caregiver, then, caring is the emotional core of at-tachment. From this core, the processes of empathy andresponsibility serve to help translate the emotion intocaregiving behavior. Whereas empathy concerns thedesire to know and understand the dependent, responsi-bility refers to the “emotional intention to help the othermeet the other’s needs” (this issue). Bell and Richardobserve that their usage of the termresponsibilitybearssome similarity to Erikson’s concept ofgenerativity.But generativity is broader and more diffuse, they pointout. According to Bell and Richard, Erikson viewed
118
COMMENTARIES
generativity as a “nondyadic need or concern that is in-ternal to the caregiver and by which the caregiver givesto future society” (this issue). Indeed, Erikson empha-sizedthatgenerativitycouldbeexpressedthroughmanydifferent channels outside of parenting—from teachingand mentoring youth to making important artistic, sci-entific, or political contributions to society (McAdams,Hart, & Maruna, 1998). By contrast, Bell and Richardsee responsibilityasa “dyadic intentiondirected towarda specific dependent growing out of feeling toward thatdependent” (this issue). Their conceptualization in thisregard is consistent with theoretical trends in develop-mental and evolutionary psychology and in cognitivescience today that underscore the domain specificity ofmany psychological functions (e.g., Pinker, 1997). Car-ing is specific to and comes directly out of a caregivingrelationship with a particular, concrete dependent:“What theparentdoes is to feel,andwhat theparent feelsis that this child has looked at this parent, that she haslooked at me” (this issue).
Bell and Richard may be right in urging us to distin-guish between the intense caring-empathy-responsibil-ity complex a mother or father feels toward her or hisownoffspringontheonehandandprovidingcare for thenext generation in a broader sense on the other. My ownexperiences as a father convince me, on an intuitivelevel, that I will never and can never feel as strongly car-ing toward any object or person as I have felt toward myown children. It may make good evolutionary sense,furthermore, thatapsychologicalmechanismdedicatedto protecting and assuring the well-being of one’s ownoffspring, thecarriersofone’sownselfishgenes,shouldbespecificallydesigned to fulfill thiscrucial task, ratherthan derivative of a more general set of all-purposemechanisms (Tooby & Cosmides, 1992). Having saidall this, Iamstill struckbythefact thatEriksonexplicitlyidentified care as the signal virtue of adult generativity.As parents, yes, but also in their roles as instructors andsupervisors, community leaders, Sunday school teach-ers, den mothers, Little League coaches, and the like,generative adults exhibit strong caring aimed towardyouth,caring that isenhancedbyempathy forothersandresponsibility to be of some good use to one’s neighbor-hood, community, people, or society, extending into thefuture. Generativity may not spring directly out ofcaregiving, and caregiving may involve different psy-chological mechanisms than, say, serving on the localschool board. But the two concepts would appear toshare more than a phenotypic similarity.
I would like to propose that what Bell and Richard soperceptively describe as the caregiving aspect of the at-tachment bond may serve as an emotional–cogni-tive–behavioral prototype of what full adultgenerativitycanbeandmaybe.Ourresearch into the lifestoriesofhighlygenerativeadultssuggests thatmenandwomen who have distinguished themselves for theirstrong and caring commitments to the next generation
typically construct narrative identities in which empa-thy for others, especially those who are weaker or whosuffer in some way, emerges clearly in early life-storyscenes (McAdams, Diamond, de St. Aubin, &Mansfield, 1997; see also Colby & Damon, 1992).Highly generative adults show an expanded radius ofcare, and they envision their futures in terms oflong-term commitments to the welfare of others (Peter-son & Klohnen, 1995). Adults who score high on pa-per-and-pencilmeasuresofgenerativitymayormaynotbe parents, but in their societal engagements they con-sistently adopt the attitude of the hopeful caregiver,looking to the future with anticipation that those thingsthey care for will grow and flourish. Undergirding theirhopefulness is what Erikson (1963) identified as a “be-lief in the species” (p. 267), a faith in the ultimateworthwhileness of the human enterprise. Although wedo not know where such a belief comes from, the beliefitself seems to suggest an expectation that those otherswhoare thepossibleobjectsofone’scarecanbeandwillbe responsive to one’s generative efforts. In the samemanner that I, the caregiver, begin to experience theemotion of caring once I see or imagine my offspring’sresponsiveness to me, so may some highly generativeadults be motivated to care for the next generation, andto commit themselves to a wide range of prosocial en-gagements, once they trust and believe deeply that thepotential beneficiaries of their care will indeed be re-sponsive to them.
In its fullest manifestations, then, especially highlevels of generativity may be experienced by some ma-ture adults as an extension of the caregiving complexto future generations and the appropriation of a widerange of activities into this caregiving program. It is nodoubt true that the many different commitments andendeavors that we may roughly group under the cate-gory of generativity may have very different originsand functions. Thus, one’s parenting may draw on dif-ferent psychological mechanisms than one’s volunteeractivities or civic obligations. But for many highlygenerative adults, these different involvements eventu-ally get organized into a generativity script(McAdams, 1993) for life, which imports the samekinds of feelings and attitudes that Bell and Richard as-cribe to caregiving in the attachment relationship. Ithink more needs to be made, therefore, of the (deep)conceptual linkages between caregiver–infant attach-ment and generativity in mature adulthood. And lessshould be made of the (superficial) similarities thathave been observed between caregiver–infant attach-ment and intimate relationships among young adults.
Note
Dan P. McAdams, Program in Human Develop-ment and Social Policy, Northwestern University,
Bowlby, J. (1969).Attachment and loss: Vol. 1. Attachment. NewYork: Basic Books.
Bowlby, J. (1988).A secure base. New York: Basic Books.Buber, M. (1970).I and thou. New York: Scribner’s.Colby, A., & Damon, W. (1992).Some do care: Contemporary lives
of moral commitment. New York: Free Press.Erikson,E.H.(1963).Childhoodandsociety(2nded.).NewYork:Norton.Hart, D. (1988). The adolescent self-concept in social context. In
D. K. Lapsley & F. C. Power (Eds.),Self, ego, and identity:Integrative approaches (pp. 71–90). New York:Springer–Verlag.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as anattachment process.Journal of Personality and Social Psychol-ogy, 59,270–280.
McAdams, D. P. (1989).Intimacy: The need to be close. New York:Doubleday.
McAdams, D. P. (1993).The stories we live by: Personal myths andthe making of the self. New York: Guilford.
McAdams, D. P., Diamond, A., de St. Aubin, E., & Mansfield, E.(1997). Stories of commitment: The psychosocial constructionof generative lives.Journal of Personality and Social Psychol-ogy, 72,678–694.
McAdams, D. P., Hart, H. M., & Maruna, S. (1998). The anatomy ofgenerativity. In D. P. McAdams & E. de St. Aubin (Eds.),Generativity and adult development: How and why we care forthe next generation(pp. 7–43). Washington, DC: AmericanPsychological Association Press.
Peterson, B. E., & Klohnen, E. C. (1995). The realization ofgenerativity in two samples of women at midlife.Psychologyand Aging, 10,20–30.
Pinker, S. (1997).How the mind works. New York: Norton.Prager, K. J. (1995).The psychology of intimacy. New York:
Guilford Press.Reis, H. T., & Patrick, B. C. (1996). Attachment and intimacy: Com-
ponent processes. In E. T. Higgins & A. W. Kruglanski (Eds.),Social psychology: Handbook of basic principles(pp.523–563). New York: Guilford.
Selman, R. L. (1980).The growth of interpersonal understanding.New York: Academic.
Stern, D. (1985).The interpersonal world of the infant: A view frompsychoanalysis and developmental psychology. New York: Ba-sic Books.
Sternberg, R. J., & Grajek, S. (1984). The nature of love.Journal ofPersonality and Social Psychology, 47,312–329.
Sullivan, H. S. (1953).The interpersonal theory of psychiatry. NewYork: Norton.
Tooby, J., & Cosmides, L. (1992). The psychological foundation ofculture. In J. H. Barkow, L. Cosmides, & J. Tooby (Eds.),Theadapted mind: Evolutionary psychology and the generation ofculture(pp. 19–136). New York: Oxford University Press.
Caregiving, Attachment, and Relationships
Harry T. ReisDepartment of Clinical and Social Sciences in Psychology
University of Rochester
Bell and Richard do the ever-expanding attach-ment area a great service by proposing thatcaregiving should occupy a more central role in thefield’s attention. Caregiving is an intrinsically impor-tant process that warrants careful thought and empiri-cal scrutiny, especially as attachment researchers ex-pand their purview beyond parents and children toconsider the role of attachment in more symmetricaladult relationships, such as may occur betweenspouses and adult siblings. At the same time, how-ever, Bell and Richard do attachment theory a notabledisservice by misrepresenting the field’s (and in someinstances, Bowlby’s) theorizing on several key con-ceptual issues. I begin this commentary by pointingout some of these misconstruals not merely to set therecord straight, but also to highlight ways in whichBell and Richard’s position is actually much morecongruous with prevailing views of the attachmentprocess than may be apparent. I hope these clarifica-tions and comments facilitate the endeavors of re-searchers who wish to study caregiving with attach-ment-compatible models.
Bell and Richard repeatedly fault attachment the-ory for relegating emotion to a nonessential functionas the by-product of attachment’s core mechanisms;for example, “this view of emotion treats the child’semotions of security or distress as signals from the at-tachment behavioral control system but not as causalparts of that system” (this issue). Although not an un-reasonable literal interpretation of some of Bowlby’searly writing (as discussed later), this is an inappro-priate characterization of contemporary attachmenttheory, and even, for that matter, of Bowlby’s deepertheoretical meaning. Bowlby described the relation-ships most likely to involve attachment as “shotthrough with strong emotion” (1988, p. 80) and as in-volving “many of the most intense of all human emo-tions” (1979, p. 69). Indeed, a lecture he delivered in1986 included a section headed “Emotionally medi-ated bonds and mental health” (1988, pp. 160–162),in which the attachment-ethological approach to“emotionally significant” relationships and mentalhealth is introduced. (Note his express use of the termmediated,which implies a causal role.) Moreover,
120
COMMENTARIES
most students of Bowlby’s writing recognize thatemotions like fear, anxiety, sadness, despair, love,contentment, relief, and joy are central to the behav-iors and phenomena he sought to explain: As he com-mented, emotions are “the prizes and penalties se-lected during evolution to guide us during ouractivities” (Bowlby, 1988, p. 81).
Admittedly, in his early writing, Bowlby(1969/1982) did restrict the role of emotion somewhatby proposing that “feeling appraisals” provide evi-dence of, rather than directly cause, the activation ofthe attachment system when threatening situations oc-cur (and hence these appraisals allow both child andcaregiver to monitor the attachment system’s activa-tion). Among several reasons, Bowlby based this limi-tation on the fact that feeling appraisals may benecessary but not sufficient to produce attachment be-havior. (Readers wishing to understand this subtle dis-tinction may refer to pp. 116–123 of the first volume ofBowlby’s, 1969/1982, trilogy). Instead, he hypothe-sized a seemingly more cognitive control system de-signed to return the child to a “set goal” of proximitywith the caregiver. (The oft-noted set goal of “felt se-curity” was added later by Sroufe and Waters, 1977).
However, it is important to recognize thatBowlby’s view of what is and is not an emotion waslimited to a large extent by the prevailing concep-tions of emotion common in his era. In the 1950s and1960s, psychology generally conceived of emotionas the subjective feeling experienced and interpretedby the individual. Thus, assigning emotion to a re-flective rather than causal role, as Bowlby did,seems reasonable, inasmuch as subjective feelingsrepresented in conscious awareness play a limitedcausal role in emotional processing and emotionalphenomena (see Ekman and Davidson, 1994, for ex-tensive discussion of this issue). But few scholars to-day subscribe to this definition ofemotion.Contemporary definitions tend to describeemotionas an automatic response to personally significantenvironmental events, involving, among other keyelements, redirection of mental attention towardthose events and increments in behavioral actionreadiness (e.g., Ekman & Davidson, 1994; Frijda,1986). Subjective interpretations are one output ofthis process, not the driving force. Although I cannotdescribe this model of emotion within the space lim-its of this comment, it is important to realize that thehomeostatic control system postulated byBowlby—and especially the role it assigns to ap-praisals of potentially dangerous environmentalevents—is very much consistent with this more con-temporary view of emotion (see, for example,Cassidy’s, 1999, synopsis of attachment theory). Infact, one may even view Bowlby’s seminal con-trol-system theorizing as having contributed to thedevelopment of modern emotion theories. Thus, in
short, appraisals of the implications of an environ-mental event for personal (or a dependent’s)well-being are at the heart of what is today under-stood as emotion, and that same appraisal process, interms of the implications of environmental eventsfor the safety of one’s dependents, is at the heart ofBowlby’s description of how the attachment systemoperates.
Historical accuracy aside, there may be little point todebating what Bowlby wrote or meant—after all, thethird and final volume of his trilogy was written morethan 20 years ago. Bowlby was far too dedicated andgenerative a scholar to have remained unaffected by thetremendous accumulation of knowledge about attach-ment, not to mention the striking and highly relevant ad-vances in relationship science, evolutionary science andethology, neuroscience, and cognitive psychology. In-stead, it may be more useful to examine the “state of theart” in contemporary attachment theory, for example, asportrayed in the newly publishedHandbook of Attach-ment: Theory, Research, and Clinical Applications(Cassidy & Shaver, 1999). Plain throughout that excep-tionally comprehensive and deeply insightful volume isthe fact that emotion, emotion regulation, and emotionalcommunication are neither secondary or incidental, norare they by-products of attachment; as attachment iscurrently conceptualized, investigated, and applied,emotion processes are a central core component.
But Bell and Richard are correct in noting that thecaregiving side of the attachment bond has receivedcomparatively scant attention (as Bowlby himselfnoted in a 1983 lecture; Bowlby, 1988, p. 82). Theirclarion call makes accuracy about the role of emotionin attachment theory all the more important. If theirmodel of the caregiving process is to serve as a guidingframework for new research, and if subsequent re-search is to be integrated with the existing attachmentliterature, then it must accurately reflect what is knownabout attachment processes. Unfortunately, on thisscore Bell and Richard fall short in several other in-stances. For example, their characterization of attach-ment activation as an on-off dichotomy rather than acontinuous “degree of activation” function reflects anoutmoded idea that even Bowlby rejected for bothlogical1 and theoretical reasons (see Cassidy, 1999;Main, 1999, p. 858). And it seems odd to assert that re-search on caregiving has “ignored the conflicts,choices, and defenses of caregivers” (this issue), aswell as dynamic motivational concepts more broadly,in light of research programs such as that of Main andher colleagues (Main, 1991), which many researchersview as paradigmatic.
121
COMMENTARIES
1That is, if the system were truly switched off, how would care-
givers notice the existence of circumstances with potential danger totheir dependents?
These reservations aside, Bell and Richard’s con-nection model offers a reasonable and much-neededstep in a vital direction, and it is to be hoped that the re-search community takes up their charge. In the remain-der of my commentary, I will focus on aspects of theirmodel that raise questions of general importance notonly to attachment researchers, but more broadly to thefield of relationships.
Where does the “caregiving behavioral controlsystem,” as Bell and Richard propose it, reside amongthe family of processes describing the provision ofhelp, support, and nurturance to others? As they men-tion, the field has struggled with the question of “whatmakes a close relationship an attachment relation-ship?” (this issue), and many researchers view attach-ment as a highly specialized bond, focused on the pro-cess of security-seeking and safety maintenance. Yetthe connection model is intended to apply to “the fullrange of relationships” (this issue), suggesting an ap-parent asymmetry in breadth of application. This dis-tinction is not theoretical hair splitting. Parentalcaregiving—overwhelmingly the domain in whichthe specialized view of attachment has been investi-gated—may express evolutionary processes that fos-ter the survival of one’s genes, which may not be rele-vant in other relationships (e.g., with a spouse or withan aging parent.) Nevertheless, caregiving is clearlyimportant across diverse relationships, many ofwhich seem unlikely to qualify as attachment rela-tionships. To what extent are common mechanismsimplicated? A further complication is that althoughparental caregiving tends to be relatively one-sided,mutuality is the norm in adult relationships, with part-ners alternating in the roles of caregiver and recipient,depending on current needs. This includes adult at-tachment relationships, such as may occur betweenspouses, siblings, and best friends. Because the con-nection model is meant to encompass both classes ofcaregiving, its mechanisms will need to be relativelygeneral, allowing for eventual integration with re-lated processes such as intimacy, commitment, andsocial support (Reis & Collins, in press). Also, themodel must be able to account for feedback betweenthese roles, as when one’s experiences as the recipientof care by a particular partner influence subsequentcare of that same partner. Presently, it is not clear howthe model would do this. On the other hand,caregiving limited to genetic dependents can proba-bly be explained by a relatively more compact set ofprinciples.
For reasons too extensive to detail presently, I be-lieve that the broader conceptualization of caregiving(and, for that matter, attachment) may have greaterheuristic value. Suffice it to say that because humansevolved in small, interdependent living groups, evo-lutionary advantage was conferred by processes thatfacilitated inclusion in close relationships and small
groups. Propensities favoring cooperative social par-ticipation thereby became an intrinsic component ofhuman nature2 (Baumeister & Leary, 1995; Brewer &Caporael, 1990). Providing care for significant oth-ers—and not just offspring—is certainly one of thesepropensities, as relationship theorists have demon-strated. For example, Mills and Clark (1982) distin-guished two important types of relationships, com-munal and exchange, defining the former in terms ofpartners’ responsiveness to each other’s needs. Oneof Fiske’s (1992) four elementary forms of social re-lations, which are presumed to be innate and univer-sal, is communal sharing, in which caring for signifi-cant others is a central element. And in spontaneousdescriptions of social goals and desired relationships,people almost invariably highlight care and support(Reis, 1990).
Although the present rendering does not directly ad-dress this issue, and although their position is con-cealed somewhat by the understandable emphasis onparental caregiving in attachment relationships, from afunctional perspective Bell and Richard’s modelseems compatible with this broader conceptualization.Many key characteristics of the caregiving behavioralcontrol system described by Bell and Richard are evi-dent in other forms of interpersonal caregiving (e.g.,empathy, feeling responsible for another’s welfare, af-fection for the other, attention to the other’s needs andresponsiveness to those needs, expressed intentions tosupport the other even when the system is dormant,possessing attitudes and beliefs that underliecaregiving behaviors). If so, it may be useful to con-strue caregiving in attachment relationships as the ap-plication in a relatively intense and contextuallyfocused way of processes operating in all relationshipsthat involve helping and support.
This issue is fundamental to understanding what arelationship is and whether relationships are betterconceptualized in terms of general interpersonal pro-cesses or whether the unique properties of each type ofrelationship merit independent investigation and a dis-tinct set of explanatory principles. (The latter orienta-tion may be seen in the development of separate andhighly specialized literatures describing particular re-lationships such as attachment, marriage, and healthcare, and those between siblings or between workersand their supervisors.) The recent emergence of “do-main-specific” approaches, which argue for the exis-tence of discrete, systematically organized modules ofknowledge and regulatory processes corresponding tothe major social tasks faced in our evolutionary his-tory, suggests a middle ground. Bugental (in press)
122
COMMENTARIES
2That is, individuals who were included in social groups tended to
be advantaged in survival and reproductive opportunities. Of course,many animal species, especially primates, display similar social in-clusion propensities, as deWaal’s (1996) fascinating analysis shows.
nominates five such domains: attachment, dominancehierarchies, coalition formation, reciprocity, and mat-ing. Domains are distinguished by their defining pro-cesses. For example, the attachment domainincorporates mechanisms that maintain safety in theface of possible threat, whereas the reciprocity domaincomprises mechanisms that maximize joint outcomesamong functional equals (including support seekingand provision). Where would the caregiving system fitwithin such a framework? At times, it seemed to methat Bell and Richard sought to move the field toward arelatively general process that may conceivably oper-ate in most domains (perhaps excepting dominanceregulation); at other times, their purview seemed rathermore limited, to the sort of processes that pertain toaffectively close relationships. Settling this issue is notonly a question of deciding to which relationships thismodel of caregiving applies; it also requires specifyingwhich forms of helpful behavior toward others are andare not subsumed by the model.
Just how important is caregiving to relationships andhow useful is it to theorize about caregiving with theconstructs that Bell and Richard propose? That remainsto be shown, of course. Bell and Richard facilitate futurework in several important respects. First, by suggestingthat motives and emotions are just as central to under-standing caregiving as information processing and so-cial competencies are—concepts that have dominatedthe literature for years—they make possible an under-standing that captures the caregivers’ purposive goal-di-rected activity and that dovetails with today’s rapidexpansion in emotion theory and research. Second, byproposing their model in behavioral control systemterms, they offer a thoughtful, dynamic perspective withconsiderable potential for linkages with related pro-cesses. And third, by highlighting the relative paucity ofresearch on caregiving, they remind us that attachmentis, after all—and notwithstanding the individualistic fo-cus of so much research in this area—a relationship be-tween interacting, interdependent individuals.
Note
Harry T. Reis, Department of Clinical and SocialSciences in Psychology, University of Rochester,
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desirefor interpersonal attachments as a fundamental human motiva-tion. Psychological Bulletin, 117,497–529.
Bowlby, J. (1969/1982).Attachment and loss: Vol. 1. Attachment(2nd ed.). New York: Basic Books.
Bowlby, J. (1979).The making and breaking of affectional bonds.London: Tavistock.
Bowlby, J. (1988).A secure base. New York: Basic Books.Brewer, M. B., & Caporael, L. R. (1990). Selfish genes vs. selfish
people: Sociobiology as origin myth.Motivation and Emotion,14,237–243.
Bugental, D. B. (in press). Acquisition of the algorithms of social life:A domain-based approach.Psychological Bulletin.
Cassidy, J. (1999). The nature of the child’s ties. In J. Cassidy & P. R.Shaver (Eds.),Handbook of attachment: Theory, research, andclinical applications(pp. 3–20). New York: Guilford.
Cassidy, J., & Shaver, P. R. (Eds.). (1999).Handbook of attachment:Theory, research, and clinical applications. New York:Guilford.
deWaal, F. (1996).Good natured: The origins of right and wrong in hu-mansandotheranimals.Cambridge,MA:HarvardUniversityPress.
Ekman, P., & Davidson, R. J. (Eds.). (1994).The nature of emotion.New York: Oxford University Press.
Fiske, A. P. (1992). The four elementary forms of sociality: Frame-work for a unified theory of social relations.Psychological Re-view, 99,689–723.
Frijda, N. H. (1986).The emotions. Cambridge, England: CambridgeUniversity Press.
Main, M. (1991). Metacognitive knowledge, metacognitive monitor-ing, and singular (coherent) vs. multiple (incoherent) model ofattachment. In C. M. Parkes, J. Stevenson-Hinde, & P. Marris(Eds.),Attachment across the life cycle(pp. 127–159). London:Tavistock/Routledge.
Main, M. (1999). Attachment theory: Eighteen points with sugges-tions for future studies. In J. Cassidy & P. R. Shaver (Eds.),Handbook of attachment: Theory, research, and clinical appli-cations(pp. 845–887). New York: Guilford.
Mills, J., & Clark, M. S. (1982). Communal and exchange relation-ships. In L. Wheeler (Ed.),Review of personality and social psy-chology: Vol. 3(pp. 121–144). Beverly Hills, CA: Sage.
Reis, H. T. (1990). The role of intimacy in interpersonal relations.Journal of Social and Clinical Psychology, 9,15–30.
Reis, H. T., & Collins, N. (in press). Assessing relationship propertiesand interactions bearing on social support. In S. Cohen, B.Gottlieb, & L. Underwood (Eds.),Social support: A guidebookfor research, measurement, and intervention. New York: Ox-ford University Press.
Sroufe, L. A., & Waters, E. (1977). Attachment as an organizationalconstruct.Child Development, 48,1184–1199.
![Parental Caregiving and Child Externalizing Behavior in ...or school), and family characteristics (e.g., socioeconomic sta-tus [SES]), that may influence the association between care-giving](https://static.documents.pub/doc/80x56/606858c9b39b0d25de6c4560/parental-caregiving-and-child-externalizing-behavior-in-or-school-and-family.jpg)
