American Journal of Orthopsychiatry, 70(1), January 2000© 2000 American Orthopsychiatric Association, Inc.

Emotional Availability: Conceptualizationand Research Findings

Zeynep Biringen, Ph.D.

The emotional availability construct (based on observations of parent-child interactions)ivasfirst reconceptualized for research in 1991 as a way to describe the quality of parent-child interactions. Since then, there has been considerable refinement of the construct. EArefers to several parental dimensions (sensitivity, structuring, nonintrusiveness,nonhostility) and tivo child dimensions (responsiveness to parent and involvement ofparent). Tire EA empirical link with attachment and parent-child relationship are revieivedand avenues for future research are suggested.

A bout two decades ago, Emde and Easter-brooks (1985) theorized that emotion islikely to be a sensitive barometer of the re-

lationship between a parent and child. They statedthat "emotional availability will, therefore, refer tothe degree to which each partner expresses emo-tions and is responsive to the emotions of the oth-er" (p. 80). Emde and his colleagues (Emde, 1980,1983; Emde & Easterbrooks, 1985) also positedthat affective attunement to a wide range of emo-tions—negative as well as positive—is an impor-tant facet of emotional availability. Further, it isnot simply physical availability but emotional avail-ability of the parent that promotes infant's self- andemotional expression (Sorce & Emde, 1981).

Mahler, Pine, and Bergman (1975) also theo-rized about emotional availability, using the termto describe a supportive maternal presence in thecontext of the child's exploratory forays and prac-tice of autonomy. In their view, the mother's "quietsupportiveness" signals encouragement and accep-tance of such explorations, and of the child's re-turns for "emotional refueling." The mother facili-tates a child's explorations, and her emotionalavailability provides a secure base for the child.Thus, in early writings on emotional availability,emotion communication was given a preeminentrole in the development of healthy adaptation.

From a different perspective, attachment theo-rists Bowlby and Ainsworth emphasized the im-portance of maternal sensitivity for the develop-ment and maintenance of a secure attachment rela-tionship (Aim-worth, Blehar, Waters, & Wall, 1978;Bowlby, 1969, 1973). In this traditional conceptu-alization, geared toward infants, the construct re-ferred to the mother's clarity of perception aboutthe infant's signals and communications, and herprompt responsiveness to them. As judged by thewealth of studies on maternal sensitivity in both at-tachment and affective-interaction areas (Biringen,1991; van IJzendoorn, 1995), the construct hasproved to be a telling aspect of the parent-child re-lationship.

Attachment theory has also emphasized the sig-nificance of "secure base" behavior to understandthe quality of the parent-child relationship (Ains-worth et al, 1978). In healthy parent-child interac-tions, the child is enabled to explore independentlyand then connect with the parent for interaction.Such moving away from, then back toward the par-ent indicates that the parent is being used as a se-cure base or haven in the relationship; movingaway is possible because the child has confidencein the parent's availability, moving toward becausethe child is sure of the parent's acceptance andwelcome. In attachment theory, the constructs of

A revised version of a paper submitted to the journal in ]uly 1998. The author is at the Department of Human Development andFamily Studies, Colorado State University, Fort Collins.

104

Zeynep Biringen

sensitivity and secure base do not emphasize therole of emotions in the parent-child relationship.

EMOTIONAL AVAILABILITY FRAMEWORKThe foundation for the conceptualization of

emotional availability that is offered here consistsof an integration of attachment (Ainsworth et al.,1978) and emotional availability perspectives(Emde, 1980; Mahler et al., 1975). Emde and Mah-ler and their colleagues emphasized the emotionaltone of interactions as a barometer of the par-ent/child relationship quality; however, they did somainly for the clinical observer.

In contrast, Biringen and Robinson (1991) of-fered a theoretical conceptualization of emotionalavailability (EA) geared specifically to research. Inthis first published account of EA, they describedthe importance of maternal sensitivity, structuring,nonintrusiveness, child responsiveness, and childinvolvement for understanding the quality of theparent-child relationship. They further describedthe historical underpinnings of these constructsand offered the EA scales for assessing them in re-search. Since then, the EA conceptualization hasbeen refined and has come to include nonhostility,as well as distinct dimensions for structuring andnonintrusiveness. Separate versions of the scaleshave also been created for infancy/early childhoodand for middle childhood. In the present article,the EA conceptualization, including four maternaldimensions and two child dimensions, is describedand the findings of research using the EA scales isreviewed.

Parental EAThe parental side of the EA concept encom-

passes parental sensitivity, structuring, nonintru-siveness, and nonhostility. The concept empha-sizes the emotional features—both parental emo-tional signaling and parental understanding of thechild's signaling—as the sine qua non of under-standing the quality and health of parent-child in-teractions.

Assessment of parental EA is based on certainprecepts. Each dimension is judged in context. Thejudgement is holistic and clinically sensitive, notfounded on counts of discrete types of behavior;for example, amount of parental smiling is less in-dicative of sensitivity than is a generally calm,contented, and relaxed emotional presence. EA isa dyadic construct; although parent and child as-pects are viewed in terms of separate dimensions,

205

it is the interaction or relationship between the twothat characterizes EA—neither parent nor childcan "look good" without taking the interactivepartner into account. Thus, a parent overwhelmedby a child's lack of clarity in emotional signaling(e.g., parents of children with autism or Down'sSyndrome) may be viewed as less emotionallyavailable in this system, because the construct isdyadic, not individual. On the other hand, a parentwho copes well with such challenges to dyadicemotional communication should not be viewed asindividually less emotionally available merely be-cause the child is, for example, autistic.

Parental SensitivityThe dimension of parental sensitivity in EA, in-

spired by Mary Ainsworth's conceptualization ofsensitivity (Ainsworth et al, 1978), is based on aclinically sensitive view of the relationship; ittakes contextual cues into account and emphasizesthe importance of such qualities as clarity of per-ception and prompt responsiveness vis-a-vis thechild's signals and communications, awareness oftiming, and flexibility.

The EA view of sensitivity, however, is broader,in that it also emphasizes affective interactions andnegotiation of conflict and dyssynchronous inter-actions. It views a relaxed climate with respect tointeractional conflict and dyssynchronies, includ-ing the successful repair of such situations (Birin-gen, Emde, & Pipp-Siegel, 1997; Tronick & Cohn,1989), as a significant aspect of sensitivity. Themost critically important aspect of EA sensitivity,though, is the role of emotion (appropriate emo-tional expression and reception). How parents notonly pick up children's emotional signals but alsoemit their own is central. For example, a parentacting very warmly can be viewed as highly sensi-tive only if, in informed clinical judgment, the par-ent's affect is perceived as genuinely positive,rather than pseudo- or forced-positive, referred toas "apparently sensitive" (Biringen, 1998). Thus,sensitivity includes both physical and emotionalresponsiveness to children's physical and emo-tional signals and communications.

A key aspect of the EA concept is that parentalsensitivity, as well as the other EA qualities, can beobserved in parents of children at any age. For thisreason, a middle childhood (focusing on 5-10years) version of the EA scales (Biringen et al.,1998) has been devised, and an adaptation for ado-lescents is envisioned.

206

Parental StructuringIn the original conceptualization of EA, structur-

ing and intrusiveness were conceived as a single,curvilinear construct (Biringen & Robinson, 1991)that ranged from unstructuring to inconsistentlystructuring to optimal to overstructuring/intrusivebehavior.

In the third edition of EA (Biringen, Robinson,& Emde, 1998), structuring and intrusiveness havebeen separated. Parental structuring refers to theability of the parent to support learning and explo-ration without overwhelming the child's autonomyand in a way to which the child is receptive. It in-volves providing rules, regulations, and a frame-work for interactions. Because EA is a dyadic con-struct that takes emotional signaling and its recep-tion into account, structuring is adequate only ifthe parent's bids or attempts at support are suc-cessful. The parent can only structure in the "zoneof proximal development" (Rogoff, 1990; Vygot-sky, 1962) by attending to the child's cues, and itis the interaction, not the parent's individual or dis-crete behavior, that is taken into account. EAstructuring, therefore, is not a means of evaluatingthe cognitive-growth fostering, quality of teach-ing, or quality of stimulation offered to the child.Optimal structuring in the context of interactionsrefers to an appropriate degree of support that al-lows for the child's reactions to the support. Par-ents whose structuring is optimal seem to provideconsistent (but not excessive) clues and sugges-tions, as well as framework, rules, regulations, andexpectations for the child and for the relationship.

Parental NonintrusivenessIn contrast to parental structuring—the parent's

ability to set limits and establish rules, regulations,and a framework for interactions—parental nonin-trusiveness refers to the ability to be available tothe child without being interfering, overprotective,or overwhelming. The quality of emotionally "be-ing there" and available when needed is indicativeof nonintrusiveness. In the early years, importantaspects of nonintrusiveness include the parent'sability to be emotionally present, both verbally andnonverbally, without taking charge; to use indirectand diversionary, rather than direct, techniques tocontrol behavior and instill obedience; and to in-teract at a moderate level that neither abandons noroverpowers the child with parental initiations forcontact. As the child grows older, nonintrusivenesscomes to include the ability to listen with an emo-

Emotional Availability

tional presence rather than "filling in" or talkingfor the child, and to grant some autonomy in mak-ing important daily decisions. At all ages, whenchildren experience difficult or challenging mo-ments, nonintrusive parents are less prone to "res-cue" and more likely to allow the children to dis-cover their own solutions. Parental nonintrusive-ness has to do with patience and allowing childrento experience the world, particularly its challeng-ing aspects, with confidence that they are equippedto meet the challenge.

Parental NonhostilityNonhostility, covert or overt, refers to ways of

talking to or behaving with the child that are gen-erally patient, pleasant, and harmonious. Althoughthe definition is clearly tipped toward the positive,nonhostile parents can nevertheless be assertivewhen necessary and appropriate, express anger ina titrated and appropriately controlled fashion, andmanage aggressive impulses. With younger chil-dren, nonhostile parents are able to remain calmand suitably controlled, even in such challengingconditions as the sleep deprivation characteristicof children's earliest months. As children growolder, nonhostile parents are able to reason and ex-plain rather than "act out" their frustrations. Thus,the nonhostile parent's emotion regulation is con-text-appropriate, and takes the child into account.

When parental hostility is displayed, it may bedirected toward the child, the self, or objects in theenvironment. Children who witness a parent punch-ing walls, breaking furniture, or physically abusinga sibling, are apt to find it as hostile and frighten-ing as if the abuse were directed at them (Osofsky& Fenichel, 1996).

Child EAChildren's emotional availability to parents may

also be viewed in terms of attachment and emo-tional frameworks. It is manifest in children's af-fective interactions with parents and in their securebase behavior, a term used here much as it is inattachment theory, in that it takes into account thechild's balance between connection and auton-omy. Whereas Ainsworth's (Ainsworth etal., 1978)use of the term referred predominantly to chil-dren's physical exploration away from followedby return to the parent, here the term refers to thebalance of emotional connection and emotionalautonomy between parent and child. Connectionand autonomy may be shown through physical

Zeynep Biringen

proximity-seeking and distancing, visual contactwith the parent followed by independent activity,or topics in conversation and play. The interactionbetween parent and child is then judged in terms ofrelatedness and autonomy.

Child ResponsivenessChild responsiveness to the parent refers to chil-

dren's age- and context-appropriate ability to ex-plore on their own and to respond to the parent inan affectively available way. A balance betweenconnection and autonomy, plus emotional respon-siveness to the parent, are the best indications ofthis quality and, presumably, of good adjustmentin the context of the parent-child relationship. Inyounger children, such responsiveness might takethe form of a happy demeanor in interaction and abalanced connection with the parent. As childrengrow older, their emotional availability to the par-ent is also manifested in more symbolic ways. Forinstance, they may take the lead in creating inter-subjective relatedness (Stern, 1985), expressingjoint attention, laughing or smiling as if child andparent shared a memory of experiences, as well aspresent interactive connectedness. In narratingthemes of play, children may show clear positivity,in both nonverbal affect and the verbal domain. Inplay, they depict parental figures as kind and lov-ing, while separation experiences are resolvedwarmly and promptly. The child views the worldas safe, secure, and benevolent. To be consideredresponsive in this framework, a child would showappropriate emotional connection with the parent,and responsiveness that was balanced with auton-omy, rather than excessive.

Child Involvement of the ParentChildren's emotional availability to the parent

can also be expressed by involving the parent ininteraction. Optimally, this would entail initiatingeye contact, asking questions, narrating a story-line, or showing and demonstrating materials tothe parent in a comfortable, nonurgent, and posi-tive manner. The child would also "socially refer-ence" (Klinnert, Emde, Butterfleld, & Campos,1986) the parent in times of uncertainty, such asnovel or stressful situations. The balance of in-volving behavior and independent activities indi-cates secure base behavior. In the early years, thechild involves the parent by visually, behaviorally,or vocally "checking in" with the parent. In the old-er child, optimal involvement is manifest through

107

a balance between seeking out the parent and inde-pendent activity. In pretend activities, the childmay assign a particular role to the parent, or spacesilences skillfully enough to give the parent therole of respondent. Optimally, the level of involve-ment of the parent would be at the "golden mean,"and would not entail over- or under-involvingstyles of interaction.

RESEARCH FINDINGSTheoretical conceptualizations require empirical

verification. The studies discussed below, someconducted by the author and her colleagues, pro-vide construct validity for the concept of EA. Ex-tant and growing evidence suggests that parentaland child EA are related to attachment, as well asto other meaningful aspects of the parent-child re-lationship.

Research that examines the link with attachmentis reviewed first, followed by research examiningEA in other contexts. Studies in the first group arethose that included the Strange Situation proce-dure, which entails two separations and two re-unions (Ainsworth et al., 1978), or the Adult At-tachment Interview (AAI), which assesses parentalrepresentations of early attachment relationships inthe parent's family of origin (George, Kaplan, &Main, 1985). Information on the validity of theseconstructs may be found in Ainsworth et al.(1978), Cassidy and Berlin (1994), Waters, Vaughn,Posada, and Kondo-Ikemura (1995), and van Ij-zendoorn (1995).

Briefly, the Strange Situation assessment yieldstwo insecure and one secure categories. Insecure/avoidant involves age-inappropriate nonchalanceabout being separated from the parent, and disin-terest/avoidance when the parent returns after theseparation. Insecure/ambivalent involves exces-sive dependence and neediness about separation,combined with inconsolability when the parent re-turns. Both insecure categories represent clearstrategies of attachment organization. In contrast,the secure category involves an infant distressedby separation but easily soothed by the parent's re-turn, suggesting confidence in the parent's avail-ability. A fourth category—disorganized/disori-ented—is additionally assigned to infants showingatypical or bizarre behavior suggesting a lack ofattachment strategy organization (Main & Solo-mon, 1986, 1990).

In the AAI categories, participants are classifiedas preoccupied, dismissing, or autonomous/secure.

108

The preoccupied individual shows significant an-ger and resistance regarding attachment figures,with little attempt to resolve or integrate relatedmemories or feelings. The dismissing individualdefensively ignores issues relevant to attachmentexperiences. The autonomous/secure individualshows balanced, integrated attachment relation-ships. A fourth category, unresolved, may be as-signed to individuals showing signs of confusion,disorganization, or unresolved mourning (Main &Solomon, 1986, 1990). The AAI was developed todifferentiate adult mental representations of at-tachment in parents whose infants had been judgedto show different patterns of attachment. Empiricalevidence suggests impressive concordance be-tween the AAI and Strange-Situation classifica-tions (van IJzendoorn, 1995).

EA and AttachmentRobinson and Spieker, 1996. This study of ado-

lescent mothers and their children examined therelation between EA at age four and earlier indicesof attachment and risk. The dyads were observedin a ten-minute, laboratory playroom, free-playsession. Various types of toys—blocks, toy cars,coloring books, crayons, a Fisher-Price dollhouse—were available for use. Interrater reliability, as-sessed as Pearson correlations on previous data,was more than .80. Lower maternal sensitivity atage four was predicted if, when the infant was oneyear old, the mother displayed depressive symp-toms and had poorer scores on the Nursing ChildAssessment Teaching Scale (NCATS) (Barnard,1979) in the fostering of socioemotional growth. Ahigh level of maternal hostility at age four was pre-dicted if the mother had a history of childhood sex-ual abuse and had lower NCATS scores on foster-ing socioemotional growth. The relationship pat-tern was clearest for boys. Curiously, a secure at-tachment predicted nonoptimal involvement, usu-ally of the overinvolving type, for the boys, where-as an insecure attachment predicted nonoptimal in-volvement for the girls, suggesting that over-re-latedness for boys in a risk sample may be a pro-tective factor.

Z;'v, Sagi, Gini, Karie-Koren, and Joels, 1996.The relation between EA and attachment also ap-pears to hold in a different culture, that of Israel.Ziv and colleagues examined the relation betweenEA and attachment, and reported the highest num-ber of insecure/ambivalent babies in the literature.The session included a six-minute free-play situa-

Emotional Availability

tion with infants aged 12 months. Interrater relia-bility was more than .80. Mothers of insecure/am-bivalent infants were found to be lower in sensitiv-ity and less than optimal in structuring/intrusive-ness, while the infants were lower in responsive-ness and involvement. The higher socioeconomicstatus (SES) group appeared to be significantlymore sensitive than the lower, and the lower SESgroup significantly more hostile than the higher. Itis particularly interesting that the lower SES moth-ers of secure babies scored higher on sensitivitythan did higher SES mothers of insecure babies.Such findings indicate that even though a mini-mum level of EA was sufficient for security of at-tachment to develop, mothers from higher SESbackgrounds created a more emotionally availableclimate for the child than did those from lowerSES backgrounds. It thus seems that EA has "ex-tra" qualities that differentiate it from attachment.Maternal hostility did not differentiate the twogroups in this sample.

Ziv, Gini, Guttman, and Sagi, 1997. In a sepa-rate report on the same sample, this group exam-ined the relation between EA in ten minutes of freeplay at six months of age, six minutes of free playat 12 months, and six minutes of free play at 20months. The Strange Situation was assessed at 12months, after the free-play session. Considerablestability was evident across the three time points.Correlations ranged from .24 to .77, with the great-est stability for sensitivity and structuring/intru-siveness between six and 12 months. Further, sen-sitivity and structuring/intrusiveness at six monthspredicted child EA (both responsiveness and in-volvement) at 12 and 20 months.

Aviezar, Sagi, Joels, and Ziv, 1999. In a morerecent study by this group, kibbutz dyads were ex-amined on three components of the attachmenttransmission model: infant attachment, EA, andadult attachment representations. Security of in-fant attachment (Strange Situation classifications)and adult attachment representations (AAI) wereboth found to be related to EA, particularly mater-nal sensitivity, with secure infants and autonomousmothers more likely to display greater EA duringinteractions than insecure infants and nonau-tonomous mothers. Maternal structuring/intrusive-ness was related only to adult attachment represen-tations. In home-sleeping arrangements only, se-cure infants had mothers with greater EA (struc-turing and sensitivity), and autonomous mothershad infants who were more responsive. Such rela-

Zeynep Biringen

tions between EA and infant-parent attachment oradult attachment representations were absent forthose in the collective-sleeping arrangement of thekibbutz. These findings suggest that the relationsbetween attachment and EA are conditional on theecological context of child care. Additional find-ings from this project are available elsewhere (Ziv,Aviezar, Gini, Sagi, & Koren-Karie, in press).

Swanson, Beckwith, and Howard, in press. Inthe UCLA FOCUS Project, an intervention study,Swanson and colleagues examined the relation be-tween maternal nonintrusiveness and attachment,as assessed in the Strange Situation, in infants pre-natally exposed to drugs. The infants were ob-served in a ten-minute play situation in the labora-tory for approximately five minutes of structuredplay with toys at a table, followed by a five-minuteteaching situation in which the mother was askedto teach the infant how to use several materials. Insome cases, the infants were observed with theirfoster mothers. Interrater reliability on maternalnonintrusiveness, assessed as Pearson correlations,was more than .80. As mentioned above, this studyused the current version of EA, which separatesthe structuring and nonintrusiveness dimensions.Maternal intrusiveness was found to be related tothe disorganized form of attachment. This suggeststhat maternal intrusiveness may lead to disorgani-zation, particularly in drug-exposed infants who,typically, are easily overstimulated. However,such attribution of directionality must be modifiedby the possibility that an infant who is disorga-nized (due to neurological immaturity, cocaine useduring pregnancy, or other reasons) increases themother's anxiety so that her behavior toward thebaby may be more intrusive.

Carter, Little, and Garrity-Rokous, 1998. Thisstudy of a group of four-month-old infants adaptedthe EA scales for young infants and examined rela-tions among maternal psychopathology, EA, andattachment. The context was before and after still-face, three-minute play segments. Interrater relia-bility, assessed as intraclass correlations, rangedfrom .76 to .92 for maternal sensitivity, .88 to .93for maternal structuring/intrusiveness, .84 to .97for maternal hostility, and .75 to .78 for child re-sponsiveness. The study found that maternal de-pression in the presence of child unresponsivenesspredicted attachment at 12 months (as assessed viathe Strange Situation). Also consistent with expec-tations, EA was significantly associated with dis-crete codings of infant and maternal behavior,

109

such as infant negativity, infant positivity, motherpositivity, and mutual positivity. Analyses con-ducted with the structuring/intrusiveness scaleused as categorical qualities (optimal structuring,inconsistent structuring, or intrusive parenting),found that infants of inconsistent mothers weremore affectively negative than were infants of op-timally structuring or intrusive mothers. Dyadswith inconsistent mothers were also significantlydifferent than those with optimal mothers on thevariables of infant positivity, maternal positivity,and mutual positive expressions. Interestingly,mothers and infants showed only moderate stabil-ity across stress-free and stressful contexts (the ini-tial play and recovery play periods of the still-facesituation).

Easterbrooks, Lyons-Ruth, Biesecker, and Car-per, 1996. With a high-risk sample of very low-in-come mothers, these researchers used the StrangeSituation during infancy to assess attachment, andthe middle childhood version of the EA scales toexamine the relation between EA and attachmentat ages seven and eight. At age seven, they ob-served a five-minute child-mother reunion follow-ing an hour-long separation. At age eight, child-mother free play, as well as child-mother reunion,were observed. Interrater reliability, assessed askappa coefficients, ranged from .95 to 1.00. Di-chotomous impaired/appropriate measures werecreated for each of the EA variables, with "im-paired" indicating maladaptive scores (i.e., thosebelow the median or any above it considered mal-adaptive), and "appropriate" indicating scoresabove the median. Findings for this high-risk sam-ple were that securely attached infants and theirparents showed greater EA at child age seven thanin infancy; parents were more sensitive and opti-mal in structuring/intrusiveness, and children moreoptimally responsive and involving of their par-ents, if attachment had been secure during infancy.Interestingly, the results held mainly for sons.When examined for type of insecure attachment,data indicated that those with disorganized attach-ment patterns in infancy displayed more impairedfunctioning than did those whose patterns hadbeen avoidant (the number of insecure/ambivalentinfants was too small to include in analyses). Thus,EA in both free play and reunion episodes was pre-dicted by early attachments, as assessed in theStrange Situation.

Easterbrooks et al. (1996) also found that mater-nal depression in infancy predicted impaired EA in

noparent-child dyads at child age seven, regardless ofwhether the mothers were still displaying depres-sive symptoms. Both maternal sensitivity and childinvolvement were impaired when mothers hadhigher depressive symptoms during infancy. Whenthe presence of an insecure attachment or of mater-nal depression during infancy was combined intoa single risk score, the composite predicted impair-ments in maternal sensitivity, structuring/intru-siveness, child responsiveness, and involvement atage seven. Again, the findings were most clear forthe boys.

Easterbrooks, Biesecker, and Lyons-Ruth, 1998.These researchers examined the relation betweenatypical maternal behavior (role-confusion, with-drawal, affective communication errors, and dis-orientation) in the Strange Situation, and EA dur-ing a five-to-ten-minute child-mother reunion aftera one-hour separation at age seven. They foundthat frequency and seriousness of the atypical ma-ternal behavior during infancy was related to bothmaternal and child EA at age seven. Further infor-mation on these findings is available elsewhere(Easterbrooks, Biesecker, & Lyons-Ruth 1999).

Siri-Oyen, 1997. This is one of two studies thathave used the AAI (George et ai, 1985) in con-junction with EA scales. It investigated the relationbetween the EA and AAI in a sample of multiriskparent-toddler dyads from low-income, single-parent families in high-risk neighborhoods. Thechildren were aged 18-42 months, with an equaldistribution of boys and girls. Observation was inthe home for 30 minutes of free play. Interrater re-liability, within one-point agreement, was 92% forthe sensitivity and hostility scales, 83% for thechild responsiveness scale, and 58% for structur-ing/intrusiveness. The low reliability on structur-ing/intrusiveness means that these findings shouldbe interpreted with caution. (The maladaptive up-per boundary for structuring/intrusiveness wascombined in analyses with the lower nonoptimalscores.) Secure mothers were found to be moresensitive and optimally structuring/intrusive, andtheir children more optimally responsive and in-volving of them. Insecure parents as a group wereless sensitive (the preoccupied parent group hadthe lowest sensitivity ratings) and less optimallystructuring/intrusive, and their children less re-sponsive, than the secure parent group. Interest-ingly, maternal hostility did not significantly dif-ferentiate the secure and insecure mothers, even inthis low-income sample, perhaps because most

Emotional Availability

mothers were able to modulate their aggressiveimpulses during the relatively short videotaped in-teractions.

Biringen, Bartholomew, Brown, Donaldson, Kr-cmarik, and Lovas, in press. This second study ex-amining AAI-EA linkages used the third and mostrecent version of the EA Manual (Biringen et al,1998), employing AAI classifications and AAIcontinuous scales in analyses. Multiple regressionanalyses indicated that with the exception of non-intrusiveness and nonhostility, each of the EAscales was predicted by the AAI classificationand/or AAI scales, with the strongest predictionbeing for maternal sensitivity.

In sum, the results on EA and attachment link-age are suggestive. The EA approach, particularly,opens up the possibility of examining combina-tions of sensitivity and moderate intrusiveness orsensitivity and low responsiveness (nonconcord-ance in parental and child scores occurs in cases offoster or adoptive homes) in predicting attach-ment. The differentiation of the parental structur-ing and nonintrusiveness dimensions in the currentversion of the EA scales also opens up the possibil-ity of examining the combined and differential ef-fects of these dimensions on attachment. Given therecent debate about the modest effect sizes in pre-dicting attachment from maternal sensitivity (vanUzendoorn, 1995), the EA approach may providecontext for a more multifaceted look at the interac-tional correlates of attachment.

EA and Other Parent-Child Relationship AspectsTwo research reports have indicated intriguing

patterns of relations between EA and discreteaffect- and control-related aspects of interaction.

Robinson, Little, and Biringen, 1993. The firstof these reported on 70 mother-toddler dyads dur-ing semistructured play in their homes at both 18and 24 months. Interrater reliabilities, assessed asPearson correlations, were more than .80. Greatermaternal sensitivity was found to predict less ma-ternal negative affect and greater maternal match-ing of affects for sons, particularly at 18 months.Lower maternal structuring/intrusiveness pre-dicted more positive child affect for sons, but notfor daughters; thus, sons appeared more positivewhen the mother was less directive or involved inplay, whereas this pattern was not seen for daugh-ters. Finally, the child's creation of shared affectstates (i.e., the child's matching of maternal affects)covaried with maternal sensitivity for daughters,

Zeynep Biringen

but not for sons. Thus, in the context of greater sen-sitivity, daughters, not sons, appeared to play a moreactive role in the regulation of the relationship.

Biringen, Robinson, andEmde, 1994. In this sepa-rate report based on the same sample, the relationbetween maternal sensitivity and initiation andmaintenance of interaction was examined. Find-ings indicated that more sensitive interactions be-tween mothers and sons were generally associatedwith sons leading the flow of interactions, whilemore sensitive interactions between mothers anddaughters were associated with greater maternalcontrol of the interactional flow. More sensitive in-teractions between mothers and sons also involvedgreater mutual control and elaboration, with inter-active partners sharing in interaction maintenance.These data suggest that sensitivity in mother-daughter interactions in toddlerhood runs some-what counter to traditional notions of sensitivity,which involve maternal responsiveness to child'ssignals and communications (Ainsworth et al.,1978). Sensitivity during mother-son exchanges,however, is more consistent with tradition.

Rethazi, Landy, and Menna, 1996. This studyexamined the relation between EA and the moth-er's representations of the child, using the Work-ing Model of the Child Interview (WMCI) (Zea-nah, Benoit, & Barton, 1993). EA was rated with asample of aggressive preschoolers and their moth-ers in a ten-minute, free-play situation in the labo-ratory. Reliability, assessed as intraclass correla-tions, was found to be high for all scales: .98 formaternal sensitivity, .96 for maternal structuring/intrusiveness, .85 for maternal hostility, .84 forchild responsiveness, and .94 for child involve-ment. Maternal representations were classified asbalanced, unbalanced/disengaged, or unbalanced/distorted. "Disengaged" referred to avoidant andaffectively distant representations, and "distorted"to confused, confusing, and preoccupied represen-tations. Mothers with balanced representationswere significantly more sensitive in interactionthan were those with disengaged representations,but no difference was apparent between motherswith balanced and distorted representations. Moth-ers with distorted representations scored signifi-cantly higher in sensitivity than did those with dis-engaged representations, and were significantlymore optimally structuring/intrusive than weremothers with disengaged representations. Thus, ina low-stress, free-play context, there seemed to befew differences between mothers with balanced

Illand distorted perceptions, but both groups appar-ently differed from those with disengaged repre-sentations.

Other research, using the Parent Attachment In-terview (Bretherton, Biringen, Ridgeway, Maslin-Cole, & Sherman, 1989), has also shown a relationbetween maternal representations of the child andEA (Biringen, Matheny, & Bretherton, 1998, inpress).

Kogan and Carter, 1996. In a low-income sam-ple, these researchers examined the relation be-tween EA and four-month-old infants' reengage-ment with their mothers after the still-face situa-tion, which involves play for five minutes, presen-tation by the mother of a motionless and affectlessface for two minutes, and "reunion" play for threeminutes. EA is rated in the initial five minutes offree play, and infant reengagement during the re-union, using scales of avoidance, resistance, andattention-seeking/maintenance developed by Ko-gan and Carter. Interrater reliability, assessed asintraclass correlations, was .90 for sensitivity, .70for structuring/intrusiveness, .78 for hostility, and.94 for child responsiveness. Infants of less sensi-tive mothers were rated as more likely to displayhigh levels of avoidant and resistant reengagementbehavior, while infants of more sensitive motherswere less likely to display such behavior.

Sagi, Tirosh, Ziv, Guttman, and Lavie, 1998.Sleep patterns at six months (e.g., wakefulness,number of times awake, and the longest period ofuninterrupted sleep) predicted EA at 12 months.Similar patterns were found for the other EA di-mensions; for example, the longest period of unin-terrupted sleep at,six months predicted child re-sponsiveness at 12 months. These findings suggestthat early styles of biobehavioral regulation in themother-infant dyad may affect the development ofEA. It would be interesting to see if such relationsbetween EA and sleep-wake patterns hold for dif-ferent children in the same family.

Robinson and Little, 1994. EA in mothers' inter-actions with their twins has also been studied in thecontext of the MacArthur Longitudinal TwinStudy, in which Robinson and Little found thatmothers of dizygbtic and monozygotic twins werehighly similar in their sensitivity and structur-ing/intrusiveness with each twin; the heritabilityestimate was only modest. However, there wasvery little similarity in twins' responsiveness to andinvolvement of the mother. These findings suggestthat within the same family, children manifest clear-

112

ly different styles of relatedness to the mother, de-spite similar maternal EA. Such findings havebeen integral to the conceptualization of EA as arelational, not an individual, quality.

Zimmerman and McDonald, 1995. ObservingEA in a day-care context over time, this study of asmall middle-class sample indicated that EA ineach relationship (mother/child, day-care provider/child) was unique and not related only to the na-ture of the infant's relationships with their moth-ers; i.e., the individual caregiver's sensitivity wasassociated with the child's emotional availabilityto that caregiver. Zimmerman is now conductingextensive observations of EA in further pursuit ofthese early findings.

Biringen, Emde, Campos, and Appelbaum, 1995.This study investigated EA in relation to infant ageand achievement of upright locomotion in amiddle-class sample. Hour-long naturalistic obser-vations were conducted in the home at severalpoints between ages nine and 14 months. Interraterreliability, calculated as Pearson correlations, wasmore than .80. Earlier (but not later) walkers showedan increase in affective expression (a measure in-cluding positive affective indices and child respon-siveness to the mother) across the shift to uprightlocomotion. A dramatic increase in mothers' sensi-tivity also was evident with increasing infant age.Thus, mothers of older infants were observed to bemore sensitive than mothers of younger infants.

Pressman, Pipp-Siegel, Yoshinaga-ltano, Ku-bicek, and Emde, 1999. In this study, EA was ex-amined in a group of deaf and hard-of-hearing in-fants, and in a group of hearing infants, all livingat home. Interrater reliability, assessed as Pearsoncorrelations, was more than .80. Maternal, but notchild, EA (particularly sensitivity) was found topredict expressive language gains, both spokenand, in the case of the deaf/hard-of-hearing infants,sign. Higher maternal sensitivity predicted greaterlanguage gains. Interestingly, child EA and mater-nal EA were not significantly different in the twogroups.

DISCUSSIONThe findings on EA reported here are thought-

provoking. They suggest that the construct notonly has a role in the prediction of attachment, butalso in the prediction or correlation of aspects ofthe child's development, e.g., language gains fordeaf and hard-of-hearing infants, sleep patterns fornormally developing infants. Given the significant

Emotional Availability

and meaningful associations between EA and dis-crete affective indices of parent-child interaction,EA also appears to be a global way of summariz-ing the overall quality of the affective relationship.

Directions for Research and InterventionParental and child EA scales provide a means to

characterize the global emotional quality of theparent-child relationship. The primary use of theEA scales thus far has been for research in the ar-eas of attachment and parent-child interaction, al-though several studies, among them that of Press-man and colleagues (1998), have also examinedthe role of EA in predicting other aspects of childdevelopment.

In addition to basic research, EA is now beingused in several intervention projects as a way toassess the affective quality of the parent-child rela-tionship pre- and post-intervention. For example,Robinson, Emde, and their colleagues (Robinson,Emde, & Korfmacher, 1997; Robinson & Gloves,1996) described the use of EA in the Home Visita-tion 2000 and early Head Start programs in groupsof very low-income, ethnically diverse mothersand their infants. EA will be used to measure theimpact of these programs on the quality of theparent-child relationship.

Can interventionists change EA through prenataland early childhood home visitation or other typesof clinical or therapeutic services? The answer tothis question is important, since EA is a way ofunderstanding the attachment-relevant, affectiveexperience between a parent and child. It may bemore possible to measure EA than attachment onrepeated occasions, and to measure EA with a vari-ety of caregivers in a variety of settings. More sys-tematic investigation of the malleability of EA isneeded through early delivery of prevention, inter-vention, and clinical/therapeutic services to chil-dren and families.

ReferencesAinsworth, M.D.S., Blehar, M.C., Waters, E., & Wall, S.

(1978). Patterns of attachment: A psychological study of thestrange situation. Hillsdale, NJ: Lawrence Erlbaum..

Aviezar, O., Sagi, A., Joels, T., & Ziv, Y. (1999). Emotionalavailability and attachment representations in kibbutz in-fants and their mothers. Developmental Psychology, 35,811-821.

Barnard, K..E. (1979). Nursing Child Assessment SatelliteTeaching Manual (NCAST). Seattle: Department of Familyand Child Nursing, University of Washington.

Baumrind, D. (1971). The influence of parenting style on ado-lescent competence and substance use. Journal of EarlyAdolescence, I I , 56-95.

Zeynep Biringen

Biringen, Z. (1991). Direct observation of maternal sensitivityand dyadic interactions in the home: Relations with mater-nal thinking. Developmental Psychology, 26, 278-284.

Biringen, Z. (1998, April). Apparent sensitivity: A method-ological note. In Z. Biringen (Chair), Emotional availabil-ity, risk, and attachment. Symposium conducted at the In-ternational Conference on Infant Studies, Atlanta.

Biringen, Z., Bartholomew, S., Brown, D., Donaldson, L., Krc-marik, S., & Lovas, G. (in press). Adult Attachment Inter-view: Linkages with dimensions of emotional availabilityfor mothers and their pre-kindergarteners. Attachment andHuman Development.

Biringen, Z., Emde, R.N., Campos, J.J., & Appelbaum, M.I.(1995). Affective reorganization in the infant, the mother,and the dyad: The role of upright locomotion and its timing.Child Development, 66, 499-514.

Biringen, Z., Emde, R.N., & Pipp-Siegel, S. (1997). Dyssyn-chrony, conflict, and resolution: Positive contributions toinfant development. American Journal of Orthopsychiatry,67,4-19.

Biringen, Z., Matheny, A., & Bretherton, I. (1998, April). Ma-ternal social cognition and object relations: Connectionswith perceptions of self and relationships with one's childand family. Poster presentation at the International Confer-ence on Infant Studies, Atlanta.

Biringen, Z., Matheny, A., & Bretherton, I. (in press). Mater-nal representations of the self as parent: Connections withemotional availability. Attachment and Human Develop-ment.

Biringen, Z., & Robinson, J.L. (1991). Emotional availability:A reconceptualization for research, American Journal ofOrthopsychiatry, 61, 258-271.

Biringen, Z., Robinson, J.L., & Emde, R.N. (1994). Maternalsensitivity in the second year: Gender-based relations in thedyadic balance of control. American Journal of Orthopsy-chiatry, 64, 78-90.

Biringen, Z., Robinson, J.L., & Emde, R.N. (1998). The Emo-tional Availability scale (3rd ed.). Fort Collins: Departmentot Human Development and Family Studies, Colorado StateUniversity.

Bowlby, J., (1969). Attachment and loss: Vol. 1. Attachment.New York: Basic Books.

Bowlby, J. (1973). Attachment and'loss: Vol. 3. Loss. NewYork: Basic Books.

Bretherton, 1., Biringen, Z., Ridgeway, D., Maslin-Cole, C., &Sherman, M. (1989). Attachment: The parental perspective.Infant Mental Health Journal, 10, 203-221.

Carter, A.S., Little, C., & Garrity-Rokous, F.E. (1998, April).Adapting the emotional availability scales for 4-month oldinfant-parent dyads: Associations with molecular codingand parental psychopathology. InZ. Biringen (Chair), Emo-tional availability, risk, and attachment. Symposium con-ducted at the International Conference on Infant Studies,Atlanta.

Cassidy, J., & Berlin, L. (1994). The insecure/ambivalent pat-tern of attachment: Theory and research. Child Develop-ment's, 971-991.

Easterbrooks, M.A., Biesecker, G., & Lyons-Ruth, K. (1998,April). Atypical maternal behavior and infant attachment aspredictors of emotional availability in middle childhood. InZ. Biringen (Chair), Emotional availability, risk and attach-ment. Symposium conducted at the International Confer-ence on Infant Studies, Atlanta.

Easterbrooks, M.A., Biesecker, G., & Lyons-Ruth, K. (1999).Infancy predictors of emotional availability in middle child-hood: The roles of attachment security and maternal de-pressive symptomatology. Manuscript submitted for publi-cation.

Easterbrooks, M. A., Lyons-Ruth, K., Biesecker, G., & Carper,

223

A. (1996, April). Infancy predictors of emotional availabil-ity in middle childhood: The role of attachment and mater-nal depression.. Paper presented at the meeting of the Inter-national Conference on Infant Studies, Providence, RI.

Emde, R.N. (1980). Emotional availability: A reciprocal re-ward system for infants and parents with implications forprevention of psychosocial disorders. In P.M. Taylor (Ed.),Parent-infant relationships. Orlando, FL: Grune & Strat-ton..

Emde, R.N. (1983). The prerepresentational self and its affec-tive core. Psychoanalytic Study of the Child, 38, 165-192.

Emde, R.N., & Easterbrooks, M.A. (1985). Assessing emo-tional availability in early development. In D.K. Franken-berg, R.N. Emde, & J.W. Sullivan (Eds.), Early identifica-tion of children at risk: An international perspective (pp.79-101). New York: Plenum Press.

George, C., Kaplan, N. & Main, M. (1985). Adult attachmentinterview. Unpublished manual, Department of Psychology,University of California, Berkeley.

Klinnert, M.D., Emde, R.N., Butterfleld, P., & Campos, J.J.(1986). Social referencing: The infant's use of emotionalsignals from a friendly adult with mother present. Develop-mental Psychology, 22, 427-432.

Kogan, N., & Carter, A.S. (1996). Mother-infant re-engagement following the still-face: The role of maternalemotional availability in infant affect regulation. Infant Be-havior and Development, 19, 359-370.

Mahler, M.S., Pine, F., & Bergman, A. (1975). The psycholog-ical birth of the human infant: Symbiosis and individuation.New York: Basic Books.

Main, M., & Solomon, J. (1986). Discovery of a new, insecure-disorganized attachment pattern. In T.B. Brazelton & M.Yogman (Eds.), Affective development in infancy (pp.95-124). Norwood, NJ: Ablex.

Main, M., & Solomon, J. (1990). Procedures for identifyinginfants as disorganized/disoriented during AinsworthStrange Situation. In M.T. Greenberg, D. Cicchetti, & E.M.Cummings (Eds.), Attachment in the preschool years (pp.121-160). Chicago: University of Chicago Press.

Osofsky, J.D., & Fenichel, E. (1996). Islands of safety: Assess-ing and treating young victims of violence. Zero to Three,16, 5-39.

Pressman, L.J., Pipp-Siegel, S., Yoshinaga-Itano, C., & Deas,A. (1999). Maternal sensitivity predicts language gain inpreschool children who are deaf and hard of hearing.' Jour-nal of Deaf Studies and Deaf Education, 4, 294-304.

Rethazi, M., Landy, S., & Menna, R. (1996, April). The rela-tionship between maternal attachment and mother-child in-teraction in a sample of aggressive preschoolers. Posterpresentation at the International Conference on Infant Stud-ies, Providence, RI.

Robinson, J.L., Emde, R.N., & Korfmacher, J. (1997). Inte-grating an emotional regulation perspective in a program ofprenatal and early childhood home visitation. Journal ofCommunity Psychology, 25, 59-75.

Robinson, J.L., & Glaves, L. (1996). Supporting emotion regu-lation and emotional availability through home visitation,Zero to Three, 17, 31-35.

Robinson, J., & Little, C. (1994). Emotional availability inmother-twin dyads: Effects on the organization of relation-ships. Psychiatry, 57, 22-31.

Robinson, J.L., Little, C., & Biringen, Z. (1993). Emotionalcommunication in mother-toddler relationships: Evidencefor early gender differentiation. Merrill-Palmer Quarterly,39,496-517.

Robinson, J., & Spieker, S. (1996). Adolescent mothers andtheir children: Early predictors of emotional availability atage four. Paper presented at the International Conference onInfant Studies, Providence, RI.

224Rogoff, B. (1990). Apprenticeship in thinking: Cognitive de-

velopment in social context. New York: Oxford UniversityPress.

Sagi, A., Tirosh, E., Ziv, Y., Guttman, S., & Lavie, P. (1998,April). Attachment and sleep patterns in the first year oflife, Paper presented at the International Conference on In-fant Studies, Atlanta.

Siri-Oyen, A. (1997, April). Multi-risk dyads: Maternal at-tachment and emotional availability. Paper presented at themeeting of the Society for Research in Child Development,Washington, D.C.

Stern, D.N. (1985). The interpersonal world of the infant. NewYork: Basic Books.

Swanson, K. (1998). Intrusive caregiving and security of at-tachment in prenatally drug exposed toddlers and their pri-mary caregivers. Unpublished dissertation, Department ofPsychology, Fuller Theological Seminary, Los Angeles.

Swanson, K.L., Beckwith, L., & Howard, J. (in press). Intru-sive caregiving and quality of attachment in prenatally drug-exposed toddlers. Attachment and Human Development.

Tronick, E.E., & Colin, J.F. (1989). Infant-mother face-to-faceinteraction: Age and gender differences in coordination.Child Development, 60, 85-92.

van Uzendoorn, M. (1995). Adult attachment representations,parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the Adult AttachmentInterview. Psychological Bulletin, 117, 387-403.

Emotional AvailabilityVygotsky, L.S. (1962). Thought and language. New York:

John Wiley.Waters, E., Vaughn, B., Posada, G., & Kondo-lkemura, K.

(Eds.). (1995). Caregiving, cultural, and cognitive perspec-tives on secure-base behavior and working models. Mono-graphs of the Society for Research in Child Development.60(2-3, Serial No. 244).

Zeanah, C.H., Benoit, D., & Barton, M. (1993). Workingmodel of the child interview. Unpubl ished manual, De-partment of Psychiatry, Tulane Univers i ty , New Or-leans.

Zimmerman, L., & McDonald, L. (1995). Emotional availabil-ity in infants' relationships with multiple caregivers. Ameri-can Journal of Orthopsychiatry, 65, 147-152.

Ziv, Y., Sagi, A., Gini, M., Karie-Koren, & Joels, T. (1996,April). Emotional availability as related to quality ofinfant-mother attachment. Paper presented at the Interna-tional Conference on Infant Studies, Providence, Rl.

Ziv, Y., Gini, M., Guttman, S., & Sagi, A. (1997, April) .Dyadic emotional availability and quality of infant-motherattachment: A three point longitudinal study. Paper pre-sented at the meeting of the Society for Research in ChildDevelopment, Washington, DC

Ziv, Y., Aviezar, O., Gini, M., Sagi, A., & Koren-Karie, N. (inpress). Emotional availability in the mother-infant dyad asrelated to the quality of infant-mother attachment relation-ship. Attachment and Human Development.

For reprints: Zeynep Biringen, Ph.D., Department of Human Development and Family Studies, Colorado State University,Gifford Building, Fort Collins, CO 80523 [e-mail: biringen@cahs.colostate.edu]