A critical exploration of the idea that parent-child attachment is contextual?

Mental health is becoming a global public health concern, “depression” especially is a significant contributor to the global crisis of many diseases and affects people in all societies around the world (WHO, 2001). With the gradual raise of the prevalence of common mental disorder (CMDs) in young people —   and adolescents (Fu, Lee, Gunnell, Lee, & Cheng, 2013), awareness of children’s health and wellbeing becomes an important topic of interest in the scientific literature, the media, and politics. Although contemporary psychologists have spawned a great amount of mental health-related research (Aspinwall & Staudinger, 2003; Seligman, 2002) — such as positive and negative psychological elements (e.g., empathy, generosity, prosocial virtues, selfishness, defensive self-enhancement, ethnocentrism) — to examine its links to wellbeing, it seems still too fragmented and specified to express positive psychology as an integrative theory. A generative and validated model, like “attachment theory” (Bowlby, 1982), is needed and be able to conceptualize both positive and negative aspects of human behaviour and experience as Mikulincer and Shaver (2005/2007) argued.

The basic concepts of attachment theory: functioning and characteristics

Bowlby (1969/1982) borrowed the concept of behavioural systems —which is based on the perspective of biologically evolved neural program to describe the process of how human beings organize behaviour in response to inevitable environmental changes and demands, in order to increase maximum chances of their survival and reproduction — to express the fundamental idea of attachment theory, like attachment, exploration, affiliation, caregiving, and sex mating. Conceptually, human beings’ inherent attachment behavioural systems are part of as an innate, functional, and goal-corrected process — including the aspect of biological function, activating triggers, primary strategy, set-goals, cognitive substrate, and interplay between attachment behavioural system and other behavioural system, which guide the choice, activation, and deactivation of behavioural cycles to attain individuals’ advantageous states of the person-environment relationship for the survival and reproduction.

According to Bowlby’s (1969/1982) presumed propositions, the “biological function” of the attachment system operates as a caregiver selector to assure that a person (especially during infancy and early childhood) can seek out a “stronger and wiser” attachment figures to maintain proximity in order to obtain protection, support, and care from them, especially during dangerous or difficult situations. Normally, when individuals are endangered by environmental threats or stressors, this situation may automatically “activate” their innate triggers of the attachment system for a need of care or protection from a selected caregiver. However, when those “activating triggers” are deactivated, or the dangers and threats are not present, people’s innate desires for care and comfort from attachment figures are terminated, they may keep pursuing their usual exploration or other activities. Theoretically, a behavioural series of proximity seeking, establishment, and maintenance to a caregiver is a person’s “primary attachment strategy” for the purpose of obtaining protection from danger, injury, or stressors in any given developmental stages. This attachment strategy can become more flexible, experienced, and context-specific with the development of goal-corrected behaviours learned from involving complex social interactions. For example, comparing to an infant’s attachment strategy (e.g., crying, eye contacting or holding), a well-cared child may develop more accurate and clear communication skills to express emotion and needs appropriately, as well as to regulate needs expression conforming to attachment figures’ preferences and role demands in order to attain protection or consolation in a given time and situation. An adult can even be able to use symbolic proximity-seeking approach (i.e., imagination or recall), instead of performing actual proximity-seeking behaviours, to consciously or unconsciously activate the mental representations of an appropriate security-providing (e.g., the memory of being comfort or calm by an attachment figure at a specific occasion) to attain a sense of safety and calm for dealing with threats or stressors.

Bowlby (1969/1982) used the term of “set-goal” to describe the ordinary cycle of how the attachment system activates and deactivates. He claimed that obtaining the sense of security is the ultimate goal of people’s attachment behaviour (especially when they encounter actual or symbolic threats, and a reliable caregiver is not available or responsive), and the attainment of “felt-security” is normally the termination of the attachment system’s activation. The cycle of attaining the goal of feeling security or being well cared for involves a series of “hard-wired” behavioural process — the experiences of threats or stress, proximity or care seeking, gaining comfort from a caregiver, feeling relieved, security and love, and returning to exploration or other activities. This process of experiencing the attainment of attachment security can provide a “prototype” or a “secure base script” to a person’s knowledge to gauge the possibility of coping with threats and obtaining the help from attachment figures, as well as to manage negative emotion in future interpersonal relationships. When a person stores enough mental representations of attachment-relevant experiences, it may enable the goal-corrected nature of attachment system for people to make the most effective adjustment in line with contextual constraints. Theoretically, this goal-corrected process requires a series of cognitive exercises, such as a person can be able to detect threatening situations and one’s own internal emotional state (e.g., stressed or felt-safety), to predict a security-provider’s response toward one’s bids for proximity-seeking, and to evaluate a chosen behaviour in a given context can be made to effectively achieve the goal of felt-security in line with contextual constraints.

In a normal attachment system, “primary attachment strategy” is one of human beings’ innate behavioural characteristics based on a hard-wired “secure-base script” of attachment system, which guide people to positively expect they can acquire the sense of felt-security through the process of the activation and deactivation of attachment behaviours. For example, people are inherent to have faith on when they encounter obstacles, danger, or stress, they can secure protection and comfort to alleviate their tension or stress, then return to other activities by pursuing proximity-seeking behaviours from a reliable and supportive attachment figure. However, when a selected caregiver repeatedly fails to meet a person’s needs in time for comfort and care, the stress and insecure feelings may continually occupy one’s state of mind, and that keeps a person’s attachment system constantly activated for the needs of being cared and protected. Time and again, when those negative interactions with an unresponsive and inadequate attachment figure prove one’s “primary” attachment strategy (i.e., proximity seeking behaviours) is unable to accomplish the goal of felt-security, the “specific” control parameter (i.e., primary attachment strategy) in one’s existing attachment system can be adjusted, and certain “secondary” attachment strategies (i.e., hyper-activation and deactivation) are likely to be activated in accordance with contextual demands (Cassidy & Kobak, 1988; Main, 1990). For instance, a person may adopt a “hyper-activation” strategies — intensifying the proximity seeking so as to claim more love, care, and attention from a selected caregiver — to deal with one’s frustrated attachment needs, which are normally caused by the inconsistent and insufficient responses, in order to obtain the comfort and stress relief —  attainment of set-goal of attachment system. A “deactivation” strategies or Bowlby’s (1969/1982) called “compulsory self-reliance” is not operated as a way to attain a dependent person’s mind-set of attachment security, but to suppress one’s needs, to weak or block one’s proximity-seeking attempts so as not to activate the attachment system. Normally, an avoidant person learns to use this “deactivation” strategy to deal with threats and distress on her or his own in order to avoid the disappointment and frustration from an attachment figure’s unavailability and carelessness.

Bowlby (1969/1982, 1973) further claimed that a primary caregiver’s responses to a dependent person’s needs not only alter the process of the attachment system in a particular interaction or short-term series of interactions, but also gradually yield more permanent and prevalent changes in attachment-system functioning. That is, through increasing storage of those significant attachment-relevant memories, it not only provides information-rich data as a guidance for a dependent person’s adjustment (i.e., goal-corrected behaviour) in time of need at a specific attachment relationship, but also benefits the formation of long-lasting “mental representations” or Bowlby (1973) called “working models” or “representational models” of important interactions with a particular caregiver. These “working” models operate as a provisional, adjustable, and context-sensitive characteristic, which allow people to predict their future interactions with a relational partner in a particular occasion, and spontaneously and unconsciously behave in accordance with context-sensitive demands. Repeated attachment-associative episodic experiences (i.e., attainment of one’s security set-goal by means of primary attachment strategy, or hyper-activation and deactivation strategies) facilitate a person’s establishment of internal working models — a generalized mental representation of external world, significant others, and the self in relation to others. According to the characteristics of these attachment working models, Ainsworth (1967) has initially conceptualized and classified them, in terms of the attributions of a child’s interactions with his or her primary caregiver (i.e., mother), into three major attachment styles — secure, insecure-anxious, and insecure-avoidant.

Normally, a person’s “prototype-like” attachment style reflects his or her most chronically accessible working models of a specific caregiver (relationship-specific attachment style) or across relational partners (global or general attachment style) at the early childhood. Children of a “secure” attachment relationship with parents, usually hold advantageous working models of successful proximity-seeking and security attainment, as a result of parents’ constantly attentive, empathetic and supportive responses toward children’s emotional needs, especially during their vulnerable moments. Children who receive this kind of secure responses from parents may consider themselves worthy of being loved by others and feel accessible to seek support and consolation from parents when they feel upset or depressed. In contrast, a child classified as “insecure-anxious” attachment style seems possess accessible attachment working models of using “hyper-activation” strategies in response to an insecure caregiver’s behaviours in order to acquire the goal of felt-security. Typically, anxious children’s maladaptive attachment behaviours as a reflection of lacking harmony and parents’ consistent responsiveness when they struggle with some difficulties and seek for emotional support (Ainsworth et al., 1978). Children of “insecure-avoidant” attachment with their primary caregivers possess a plenty of working models that are beneficial to “deactivate” their security-seeking functioning of attachment system. Normally, avoidant children have been chronically treated by a parent’s neglect, rejection, and unresponsiveness of their proximity-seeking attempts in time of need (Ainsworth et al., 1978). These insecure parent-child interactions (i.e., attachment system hyper-activation and deactivation) may contribute to children’s low value of themselves and be reluctant for seeking help or depending on others in their future social networking.

Attachment styles and psychological-related outcomes

The essentiality of attachment process can be based on the concept of emotion regulation (Shaver, Schwarts, Kirson, & O’Connor, 1987) to describe how individuals learn the way of how to deal with their negative emotions and fragility through the interactions with parents in order to achieve their attachment goals (i.e., attainment, or deactivation of the needs, of being cared and protected) (Mikulincer & Shaver, 2016). Theoretically, the ability of a person’s “emotion regulation”, developed from the parent-child interactions in the early developmental stages, can be the most influential indication for one’s future mental concern — whilst an adult’s mental illness cannot be exclusively attributable to parent-child maladaptive attachment functioning, other innate (e.g., intelligence, disposition) and external factors (e.g., life transition, traumatic events) may also account for the causes of it. For example, an anxiously attached child is likely to develop a dangerous, hopeless, and unpredictable working model of world as a result of an attachment figure’s inconsistent and unreliable responses. This type of insecure children are not only doubt of their worth and efficacy to obtain love and protection from a selected caregiver in time of need, the resulting unfulfilled needs of “felt-security” also cause them to intensify negative emotions (e.g., angry and cry) or implicitly accent their vulnerability and neediness (e.g., anxiety, fear, and shame) as a “down-regulation” strategy to keep pursuing their attachment goal (i.e., gain a protector’s attention or care). They are not only disadvantageous to learn to use well-adapted approaches to regulate “negative emotion” from maladaptive parenting, but also use “negative emotion” as a strategy to capture attachment security. Suffering from chronic emotional dysfunction, despite increasing the likelihood of capturing a caregiver’s attention and protection, children’s constantly unmanageable cognition and feelings of angry, fear, and anxiety can be transformed into mental problems (e.g., depression or anxiety disorders). Comparing to insecure-anxious people, avoidant children are normally nurtured in a neglected, unresponsive, and emotionally-distant parenting environment. Their painful interaction experiences with cool and rejecting parents encourage the development of deactivation (defensive inhibition) strategy to avoid themselves from involving in any emotional states (e.g., fear, anxiety, anger, sadness, and distress) when they face with threats or distress, because these negative emotions may induce attachment system activated (i.e., proximity seeking) and arise disappointed and depressive experiences and emotion. Unlike securely attached people’s emotional regulation strategy (e.g., communication, compromise, relationship maintenance), avoidant children are preferred to use self-reliance approach (e.g., block negative emotions, switch off any emotion-related attentions, inhibit the expression of emotion) to minimize the risks of suffering from emotional disturbance. However, keeping interpersonal distant and suppressing negative emotion can leave unresolved distress behind, and disable one’s ability to handle with unexpected life adversities, especially when they are encountering long-term and highly stressful situations. These damages of maintaining insecure people’s emotional defensive system can be transferred to psychological disorders.

In Bowlby’s proposed concept and key tenets of attachment theory (1969/1982, 1973, 1980, 1982), he especially stressed the importance of a person’s ability of “emotion regulation” and how this “learned abilities” during one’s early attachment experiences (i.e., infancy, childhood, and adolescence) render one’s later psychological well/ill-being. His concerns have also been recognized by several scholars from different academic and clinical fields (e.g., psychoanalysis, psychopathology, psychotherapy, and psychiatry) to look into the relationships between early parent-child attachment interactions and the resulting psychological-related outcomes. In a review of previous attachment-based psychopathological relevant studies (see chapter 13, Mikulincer & Shaver, 2016), it showed that attachment characteristics (i.e., insecure-anxious, and insecure-avoidant) is substantially associated with both of internalized mental problems (e.g., anxiety disorders, depression, post-traumatic stress disorder) and externalized behavioural issues (e.g., suicide tendencies, eating disorders, substance abuse, criminal behaviours, personality disorders, and dissociative disorders). That means, if people’s attachment experiences (especially during early developmental stages) are largely associated with their mental issues, it seems important to understand how the human beings’ attachment system functioning and changes across the lifespan in order to provide some appropriate and timely interventions to terminate their deteriorating attachment behavioural patterns, and boost attachment security through cognitive or behavioral treatment.

The continuity, stability and fluctuation of attachment styles

The concept of attachment dynamics

While the concern of stability and change of a person’s early attachment bonds with primary caretakers (usually parents) has been broadly explored and be considered to be relatively stable throughout the life courses (e.g., Fraley, 2002; Klohnen & Brea, 1998; Simpson, Collins, Tran & Haydon, 2007; Dykas & Cassidy, 2011), how fluctuation (i.e., state-like, trait-like or interaction effect between momentary and permanent state of mind) of one’s attachment orientations during the lifetime is still an arguable issue. Initially, Bowlby (1973) combined the Waddington’s (1957) concept of “epigenetic landscape model” (i.e., environmentally stable, environmentally liable) to emphasize “dynamic” functioning of human beings’ attachment working models. He asserted that individuals’ development of attachment patterns is simultaneously harmonized by two innate strengths — “homothetic” and “destabilizing” forces. The force of homothetic helps that ones’ early attachment working models keep on their assigned track throughout any phrase of life (e.g., infancy, childhood, adolescence, adulthood). However, the other opposite inherent force allows people to destabilize their early-formed attachment representations by continually updating attachment-related new experiences. In a sense, human beings’ attachment representations can be spontaneously operated by both processes of “assimilation” and “accommodation” — not only take in new experiences to existing mental representations, but also revise previous working models to accommodate current attachment associative experiences (Bowlby, 1969/1982).

Subsequent attachment theorists (e.g., Crowell & Waters, 2005; Fraley, 2002) further proposed a concept of “prototype” perspective to characterize how these two contradictory forces work together to form one’s attachment style at a given developmental stage. They claimed there are two separate working models — “prototype-like” and “current” working models — that concurrently function to shape a person’s “phase-specific” attachment characteristics. Specifically, a person’s current working models can be revised and updated throughout the lifespan by one’s “present” experiences deviated from his or her existing attachment-relevant beliefs and knowledge. However, prototype working model is rooted from a person’s infancy, and extends an influential impact across one’s developmental stages. Thus, when infants enter into childhood, their previous attachment schemas (i.e., “prototype” working model developed during “infancy”) may synchronically function with current working models — those live experiences of ongoing interactions with relational partners (e.g., parents, relatives, siblings, close friends) — to shape their “childhood” attachment pattern. In other words, albeit a person’s “prototype” working model plays a fundamental and prevailing role on retaining that one’s attachment styles later in life will reflect the earliest working models (i.e., infancy), prototype-like models still can flexibly incorporate incompatible attachment relative experiences happened during one’s “present” and “previous” developmental phase(s), thereby resulting in “structural/qualitative” change of stage-specific attachment schemas. For instance, when “securely” attached adolescents — who probably also have “secure” models during “infancy” and “childhood” — frequently take in experiences of being rejected or neglected by surrounding attachment figures, their existing beliefs of secure attachment can be eventually dominated by those continually conflicting experiences/memories, and therefore, taken to extremes, result in “adolescence-specific” attachment insecurity (see Mikulincer & Shaver, 2016, p.112).

Comparing to the other main stream of attachment systems functioning — the view of “continuous change” or “revisionist”, “prototype-approach” attachment operation seems to be more favoured and sensible to explain individuals’ fluctuation of attachment throughout the lifelong time and previous inconsistent research in the continuality of attachment characteristics (Mikulincer & Shaver, 2016; Fraley, 2002). According to Fraley’s (2002) meta-analysis study regarding to attachment stability from infancy to adulthood, it indicated that “prototype-perspective” process has a moderate-level variance of approximately .39 on the stability of one’s attachment orientations across one’s different developmental stages (especially till 19 years old). This result seems correspondent to previous similar research (e.g., Owens et al., 1995; Fraley & Shaver, 1999; Shaver, Belsky, & Brennan, 2000; Bartholomew & Shaver, 1998) that found around a correlation of 30% between a person’s early attachment security with parents and one’s later romantic relationships. These empirical evidences bring about more questions, like how to explain an unknown variance of up to 70% in people’s attachment stability across their life courses? What situational (e.g., interpersonal interactions, significant life events, unconscious or conscious inventions) or developmental (e.g., biological, cognitional, affectional maturity) factors can be used to account for those fluctuations? Given previous substantially empirical studies (e.g., Davila & Sargent, 2003; Marks, Trafimow, & Rice, 2014; Mikulincer & Shaver, 2007; Overall, Fletcher, & Friesen, 2003; Sibley & overall, 2007, 2008; Gillath, Selcuk, & Shaver, 2008; Cozzarelli, Karafa, Collins, & Tagler, 2003) on the change and continuity of a person’s attachment working models, we would like to briefly summarize some validated findings (i.e., attachment functions and needs across the relationships, symbolic and semantic priming, the meaning of life events) that may be able to account for fluctuations of one’s “current” internal working models across the lifespan, and how these within-person variations of mental representations link to a broader and generic schema as a hierarchical (and/or multidimensional) attachment network.

Multiple working models in relational network

Along with growth of age and developmental maturity, the major life realms and interpersonal network of a person may progressively expand in size and diversity, like transfer from very core of familial life to schooling, extracurricular activities (e.g., sport), workplaces, social networking events, etc.). This expansion and transition of life focuses can be conducive to a person’s formation of, apart from parents, a wider variety of attachment bonds with multiple figures — grandparents, older siblings, neighbours, relatives, close friends, teachers/coaches, coworkers, romantic partners and spouse — as subsidiaries for closeness and sources of security when, needed, and primary caretaker is absent (Sukys, Lisinskiene, & Tilindiene, 2015; Carr, 2012; Seibert & Kerns, 2009; Furman, 1989; Furman & Buhrmester, 1985; Weiss, 1974). For a dependent person, “attachment figures” can be someone who can serve the function of, like, proximity maintenance, safe haven and safe base, yet these figures cannot be interchangeable, or have no difference to a dependent person. Comparing to other relational partners, “parents” (especially mother) are normally one’s “relatively” preferred, chronical, crucial and influential figure in the attachment hierarchy throughout one’s developmental trajectory — this is what Bowlby (1969/1982) called “monotropy” tendency.

Previous studies have documented that the role of “principle” security providers seems possibly to be replaced by other significant figures in a given developmental stage, for example, parents are mostly like a person’s primary caregivers till late childhood, and close friends and romantic partners can become the preferred and prevailing attachment figures to adolescents and adults  (Fraley & Davis, 1997; Hazan & Zeifman, 1999; Howes & Spieker, 2008; van IJzendoorn & Sago-Schwartz, 2008). However, it does not mean that a person’s “monotropy” tendency to parents is no longer existing when grow up, it is just that parents and other preeminent figures jointly play different roles in attachment functions (i.e., proximity providing, safe haven, and secure base) and positions (i.e., principle and secondary caregivers) of one’s developmental phrases (Allen, 2008; Hazan & zeifman, 1994; Trinke & Barrholomew, 1997; Schachner, Shaver, & Gillath, 2008). For instance, school-age children may start to target peers for the source of proximity seeking, and not till late adolescence or early adulthood parents are still their principal providers of safe-haven and secure-base functions. Adults normally more prefer close friends and romantic partners, but parents, to serve as safe-haven for comfort and support. However, for the primary providers of secure-base function, they may consider romantic partners, or parents if they are not involved in a long-term romantic relationship or marriage, more than close friends. Furthermore, other theorists (e.g., Overall, Fletcher, & Friesen, 2003; Lewis, 1994; Pierce & Lydon, 2001; Trinke & Bartholomew, 1997) also claimed that individuals’ attachment-related needs (e.g., closeness, social and exploration, safety and nurturance, sexual fulfillment, etc.) may be varied across specific relationships or relational domains (e.g., familial, friendship, romantic). Like they may possess attachment expectations on being more close, dependable and exclusive in their romantic relationships, whereas more concern their needs of social and exploration within friendship domain, and higher expectations on the needs of security and nurturance across familial relationships.

How these figures providing diverse attachment functions fulfill a dependent person’s needs throughout the life courses may also make a considerable impact on the variations of within-person attachment orientations across relationships (La Guardia, Ryan, Couchman, & Deci, 2000; Lewis, 1994). In light of the conceptual links between relationships motivation theory (RMT) and attachment theory, La Guardia et al. (2000) applied the perspective of RMT to explore how a person’s satisfaction of basic psychological need with relationship-specific figures (e.g., mother, father, romantic partner, best friend) can explain the variations of a person’s attachment styles. They contended that the satisfaction of needs for relatedness seems not enough to bring about a flourishing relationship, the fulfillment of needs for autonomy and competence are also required in a given relationship. Their sequent studies majorly examined the influence of a person’s perceptions of basic needs satisfaction on attachment security across the relationships at between-person and within-person level. The results indicated the variability of needs satisfaction at “within-person” level can account for approximately twice as much variance in the attachment variables (i.e., self, other and overall security) as the between-person level. In detail, adults’ satisfaction of needs for “autonomy” can significantly predict their attachment security within specific relationships with “romantic partners”. Plus, the perceptions of attachment security, especially with “parents” and “peers”, can be predicted by both of the fulfillment of needs for autonomy and competence (after controlling the needs for relatedness). General speaking, grow-up people seem to have different attachment expectations and needs (especially for autonomy and competence) toward specific figures. The greater satisfaction for the specific need(s) they experience with the correspondent relationship-specific partners can reflect on the greater attachment security within a particular relationship. Ones’ overall attachment security can be the coverage of the variations of within-person attachment — as a function of the satisfaction of particular needs within relationship-specific figures.

Security priming-related research

Given a person’s perceived fulfillment level of attachment needs from interacting with relationship-specific figures at a given developmental stage may give rise to various attachment working models that representing particular cognitional and affectional beliefs (e.g., motives, goals, emotions) within each of the relationships, thereby resulting in the variations of within-person attachment styles, and the subsequently correspondent psychological-related outcomes. Previous numerous theorists have applied the technique of “priming” to test the validation of change of within-person attachment patterns, and its short-term and long-run effects on individuals’ psychological-related outcomes, as result of the fact that the maintenance of attachment security has considerable influences on ones’ wellbeing (see the reviews, Mikulincer & Shaver, 2007a; Gillath, Selcuk, & Shaver, 2008). To date, popular security priming methods in laboratory-based studies, like exposing participants to proximity-relevant words (e.g., love, closeness), the names or pictures of secure attachment figures, asking to recall or imagine some scenarios of being loved and cared by security providers, have been found their positively effects on people’s sense of security, and mental health-related issues (e.g., mood, aggression, compassion, altruism, eating disorders) (Mikulincer et al., 2001; Mikulincer, Hirschberger, Nachmias, & Gillath, 2001; Mikulincer & Shaver, 2007a; Gillath, Shaver, & Mikulincer, 2005; Mikulincer, Shaver, Gillath, & Nitzverg, 2005; Admoni, 2006). This is for the reason that the effects of the temporary activation of mental representations of attachment security by these subliminal priming practices (i.e., semantic, symbolic) can bring about the outcomes (e.g., the senses of calm, soothing, protective, healing) similar to a person’s actual experiences of interacting with that designed security-enhancing figure in real life.

Moreover, the similar effects of security priming on a person’s cognitive representations can also be found in the previous studies (e.g., Baccus, Baldwin, & Packer, 2004; Amdt, Schimel, Greenberg, & Pyszczynski, 2002; Schimel, Amdt, Pyszczynski, & Greenberg, 2001; Rowe & Carnelley, 2003; Rowe & Carnelly, 2007) examining the variables in association with self (e.g., self-worth, state self-esteem, defensive self-inflation, defensive self-handicapping) and relational partners (e.g., positive expectations toward a partner’s behaviors, relational strategies, long-lasting positive effects on relational beliefs). In detail, these findings are in accordant with the positions of attachment theory that a person’s regular experiences (by actual or recalled interactions) of specific attachment figure’s availability, sensitivity and responsiveness in times of need can reassure one’s positive senses of self-worth, decrease the necessity of actively using defensive strategies to protect self-esteem (Mikulincer & Shaver, 2005). The beneficial effects of security priming also reflect on bolstering one’s positive expectations (e.g., beliefs on a partner’s good qualities – loyalty, being considerable and respectful in the expression of needs), willingness of being close and seeking emotional support, to a relational partner.

Important life events

On top of previous priming-related literature boosting state/momentary attachment security of oneself through repeatedly visualized manipulations to trigger participants’ proximity-associated mental representations with relationship-specific partners, thereby improving one’s wellbeing and the relationships with significant others in the short term or long run. Other theorists (e.g., Cozzarelli, Karafa, Collins, & Tagler, 2003; Davila & Sargent, 2003; Feeney & Noller, 1992; Hammond & Fletcher, 1991) have also focused on different angles — like the effects of important life events on the fluctuation of within-person attachment. For example, a series of studies by Davila et al. have shed light on the issue of how those day-to-day basis life events affect a person’s state changes in attachment security (Davila, Burge, & Hammen, 1997; Davila, Karney, Bradbury, 1999; Davila & Cobb, 2003; Davila & Sargent, 2003). In a key study by Davila and Sargent (2003), they considered the importance of viewing a person’s state changes of attachment security as contextual or relational-specific cases, which means those “present” mind-scripts of attachment a person holds on (e.g., beliefs, attitudes, interpersonal feelings and behaviours) can be a function of, (or primed by), a variety of situational-specific factors (e.g., social circumstances, daily interpersonal experiences). They argued one of the reasons that prior research (e.g., Beckwith, Fehr, 1995; Davila et al., 1997; Scharfe & Bartholomew, 1994) has failed to constitute a valid prediction of life events to the change in attachment security over time was as a result of the inappropriate definition to “life events”. That is, most of previous researchers primarily examined the “objective” attributes (such as the type or number of life events), instead of “subjective” experiences — like the appraisal of the emotional or interpersonal factors relating to important life events. In their study, participants were not only asked to indicate the negative life events representing various life/relational domains (like regarding to school, family, friendship, romantic relationships, extracurricular activities) that have happened in the previous 24 hours, but also to rate how much the events in relation to their interpersonal and achievement loss. The results revealed that, as they expected, “interpersonal loss” (e.g., loss of emotional support, closeness or affection, friendship or companionship ones had or wanted), but achievement loss (e.g., loss of reaching goals, meeting expectations), can significantly account for the variance of attachment insecurity on a daily basis. In other words, when people perceived higher interpersonal loss related to their assigned life events, they showed higher perception of attachment insecurity on a daily basis.

Can a person’s attachment representations across relationships operate as hierarchical working models?

Abovementioned previous research on the fluctuation and stability of people’s attachment security seems in line with theoretical proposition (Bowlby, 1973) — the formations of a person’s attachment characteristics can be through actual interactions with multiple caregivers, especially those experiences (or mental representations) with parents (normally play the role of monotropy) during early developmental stage can be fairly endurable and stable across the life courses, but still open to be changed. Social-cultural factors, such as attachment experiences with various relational partners, important life events relating to different relational domains, etc., are likely to operate as crucial antecedents to temporarily change a perosn’s attachment security — and therefore form a variety of “state-like” mental reprsentations within specific relationships. The most frequent and chronical appearances in the working models within a given relationship are possiblely to govern a person’s attachment orientation with a specific partner across the relationships. Numerous literature (e.g., Kenny, Kashy, & Bolger, 1998; La Guardia et al., 2000) has documented that examining the causal relationships between people’s attachment types on psychological-related outcomes at “within-person” level seems more superior and valid than at “between-person” level. Furthermore, considering the fluctuations of a person’s working models across the relationships can be more accuarately to predict the attachment-related sequencetances (e.g., life satisfaction, the quality of social interactions, well-being) than the global (or trait-like) attachment schemas (Pierce & Lydon, 2001; Cozzarelli, Hoekstra, & Bylsma, 2000; La Guardia et al., 2000).

The issue of how the relationship-specific working models in association with global-like attachment representations has been paid much attention by several researchers (e.g., Collins & Read, 1994; Fraley & Davis, 1997; Trink & Bartholomew, 1997; Pierce & Lydon, 2001; Overall, Fletcher, & Friesen, 2003; Imamoglu & Imamoglu, 2006). For example, in the studies of Imamoglu & Imamoglu (2006) and Pierce & Lydon (2001), both revealed the similar results in the relationships between global and relationship-specific attachment working models. Like, individuals possessing higher attachment security in global level, comparing to lower security ones, were signficantly related to higher security across the relationships — family, romantic and peer/friend relationships. Nevertheless, participants did not have the same perceptions of attachment security across the relationhips as their global levels, instead, they perceived more security in family domain than in others.

Collins and Read (1994) proposed a default hierarchy of attachment working models, in

According to the concept of Bowlby’s (1969/1982) “monotropy”, and previous research on developmental changes in the hierarchy of attachment figures (Bretherton, 1995; Hazan & Zeifman, 1994; Freeman & Almond, 2010), we can briefly summarize that an individual seems able to form “context-specific” (e.g., family, school, sport clubs, extracurricular settings, workplaces) attachment bonds with different significant others. This is due to those figures (e.g., parents, older siblings, close friends, teammates, teachers, coaches, coworkers, romantic partners) can be more accessible and attainable (than parents) to satisfy “specific” attachment function (e.g., proximity, safe haven, and secure base) in time, when needed, in a given milieu and developmental stage. And empirically, when a person perceives more security, support and needs satisfaction from a specific attachment figure, the higher level of that relational partner’s predominance in the attachment hierarchy (Keren & Mayseless, 2013; Milyavskaya & Lydon, 2013). The experiences of interacting with “relationship-specific” figures can also be the sources of revising and updating one’s “current” working models at a given developmental stage. And attachment functioning in the “relational” domain is likely to operate as the perspective of “prototype-approach” working models, which includes a “prototype-like” working models with a specific attachment figure and “current” working models with other subsidiary figures.

A child’s early interacting experiences with mother forms “prototype” attachment working models and that may exert a dominant and long-term influence on one’s future experiences in other relational partners or relational domains, but it is still open to be adjusted by “current” relationship-specific experiences. The reciprocal influences between an individual’s “prototype” and “current” relationship-specific working models may contribute to the consolidation or destabilization of prototype attachment representations in a given relational domain, thereby separately forming more generic relationship-domain working models (e.g., familial, friendship, romantic) and global representations of self and others across various relational domains.

More specific, although the “prototype” attachment representations rooted in one’s very early ages can exert an implicitly and chronically influence on the choice of accessible attachment strategies (i.e., proximity seeking, hyper-activation, deactivation) for interacting with other attachment figures, theoretically the memories of the most stable and chronical relationship a person involves in with an attachment figure are also likely to become a “prototype-like” working models in a relational domain. For example, with the recurrently accumulation of relevant memories with a “securely” attached partner in a long-term relationship (usually more than one year), those supportive and responsive attachment-related experiences may increase a person’s likeliness of accessing “secure” attachment representations and that can also “inhibit” the accessibility of “insecure” attachment working models. With the frequent interactions with other relational partners, these prototype-like models can also be conducive to form excitatory and inhibitory connections with other relationship-specific attachment models in a given relational domain (e.g., friendship, romantic). When these connections are reinforced and consolidated, people can be able to form an even more abstract and generalized working models — global representations of self, others and world — as a hierarchy of attachment relevant experiences. In a way, those dominantly and dramatically episodic memories can be the typical representations of relationship-specific models, and these models become exemplars of generic relational representations, then those generalized domain-specific representations can be the abstract exemplars of global attachment schemas.

The conceptualization of “contextual” attachment representations in parent-child relationship

Grounded Bowlby’s (1973, 1969/1982) concept of attachment dynamic systems and previous findings on the issues of “within-person” (i.e., variations of attachment styles with multiple relational partners) and “between-person” (i.e., individual differences in attachment characteristics and its influences on psychological outcomes), numerous attachment-related researchers further looked into the topic of why an individual’s attachment styles change. According to Bowlby’s concept of dynamic behavioural systems, a person’s attachment behavioural system seems able to spontaneously interplay with other nonattachment behavioural system (e.g., exploration, affiliation, sexual activities) in order to facilitate one’s survival chances, personal growth and self-actualization. In other words, not all interactions between a dependent person and a primary caregiver can be called “attachment” interactions. When people are not under conditions of threat, fear, or upset, they may engage in other nonattachment activities with an assigned attachment figure. For instance, a child and a parent may go shopping, play games at home or joke each other together, these types of interactions can be more like “affiliation” behaviours that do not service so-called “attachment” functions (Bowlby, 1969/1982; Weiss, 1998), although it may be beneficial for maintenance and cementation of their affectional and behavioural harmony, thereby increasing accessibility of attainment of attachment security when needed. So what kind of “life events” can really activate people’s attachment system functioning (i.e., proximity seeking, selection of attachment strategies)?

In light of previous conceptualization of a “hierarchy” of attachment working models in relational domains and “state-like”/”trait-like attachment representations, it seems possible to regard human being’s attachment system functioning as a hierarchical and multidimensional model in which state-like episodic representations across different relationships can be converged and categorized into more generic relationship-domain schemas, and those schemas become typical characteristics of a person’s global attachment representations. However, previous attachment-relevant research seems to lack a piece of puzzle — contextual (or context-specific) working models — in this hierarchical associative attachment network. That is, a great deal of previous findings have demonstrated that the role of life events during a person’s different developmental stages (i.e., from childhood to adulthood) are associated with one’s changes of attachment styles on day-to-day basis  (e.g., Beckwith, Cohen, & Hamilton, 1999; Weinfeld, Stroufe, & Egeland, 2000; Egeland & Farber, 1984; Egeland & Stroufe, 1981; Vaughn, Egeland, Stroufe, & Waters, 1979), especially how a person’s perceptions of “interpersonal loss” related to life events, but not simply the occurrence of events or the type and number of events (Davila & Sargent, 2003).

It seems possible that a person’s appraisal of the “importance” and “meaning” of life events might be varied with their surrounding significant others and emphases in daily life at different ages. For example, “parents” are probably the major attachment figures who a person spends most of time with during school-age period (Higginson, 1985). The majority of life focuses for schoolchildren can be family time, schooling and/or other extracurricular activities (like sport), parents’ behaviours and emotional responses toward children’s encounters of negative “life events” (e.g., lost a game, fail in a test), in which they may perceive loss of parental support (e.g., rejection, neglect), may result in their temporary changes of attachment styles (i.e., “state” attachment representations), and this “state” changes can last for a relatively long term (Gillath, Selcuk, & Shaver, 2008). According to the concept of hierarchical associative memory functioning, we presume that through frequent access (actual experience or thinking of) secure/insecure attachment-related working models with a relationship-specific figure in some specific life events, these episodic memories seems possible to be classified into more generic and abstract “context-specific” attachment representations, and these “contextual” working models can be typical exemplars of “global” attachment schemas with a specific relational partner. To date, except for two studies that provided some promising notions (Lai & Carr, 2017a, 2017b), none of contemporary attachment-related researchers have awareness of this gap stratification between a person’s “state-like” and “global” working models with a specific attachment figure. According to Lai and Carr’s conceptualization of “context-specific” attachment relationship, “context” can be defined in a more complicated and broader way, but not simply as its literal meaning. That is, those negative life events causing a person’s emotional disturbance and pursuits of attachment goal may not only happen in a “tangible” place or venue (e.g., school or competition arena), but also can be found in any daily spoken or unspoken conversations, like facial, affectional and gestural expressions at any occasions. Therefore, it seems more sensible to conceptualize “contextual attachment” as more generic and typical attachment schemas of representing any temporary (state) working models related to the interactions between a person and a relationship-specific figure in one’s negative life episodes.

Can school-age pupils develop “sport-specific” and “schooling-specific” working models? 

In light of “monotropy” tendency (i.e., “parents” are likely to be the prevailing attachment figures who stand out from others) and “prototype-approach” (i.e., parent-child attachment bonds developed during infancy can last throughout the life course) of attachment systems functioning, in the present article we are more interested in illustrating the rationale of how the concept of “contextual-like” parental attachment representations can especially be applied in the contexts of sport and schooling, which are two major life domains that schoolchildren may largely involve in achievement related activities. As alluded to earlier and previous numerous studies of individual difference in the relationships between attachment patterns and psychological outcomes within a given context (e.g., sport, school),

Schooling policy and objections on children’s knowledge acquirement and performance recognitions in academic and sport achievement may likely encourage school principals, teachers, coaches, learners and parents to pursuit dysfunctional learning motivations and achievement goals (i.e., ego-involved/performance goal). This maladaptive learning atmosphere can not only thwart children’s learning outcomes and psychological need satisfaction, but also lead to parents applying pressure, becoming obsessive and inducing anxiety and stress. For example, athletic is a highly public context, which provides opportunities for interpersonal interactions, that significant others’ (e.g., parents, coaches, teammates, etc.) are likely to shape children’s experiences in sport participation through immediate and specific feedback during their routine practices and competitions (Hellstedt, 1995, Fredricks & Eccles, 2004). The motivational climates (i.e., performance goal) created by those relational partners may result in unduly high expectations on comparison and proof of self-competence, but not pursuit of self-improvement and attainment of one’s own set-goals. This need-thwarting atmosphere in sport setting may greatly induce children’s proximity-seeking behaviours, especially when they lose a game or being a “bench-warmer”, children might need more support and consolation from their parents. But parents likely become too busy with their own aggressive and competitive feelings and respond children’s emotional needs in a maladaptive parenting way (e.g., scold or neglect their needs for support).

Maladaptive parenting practices in sport field have been significantly noticed by several sport psychologists or psychiatrists, for instance, Tofler & Butterbaugh (2005a) and Tofler, Knapp, & Lardon (2005b) noted that sport and schooling are the contexts which particularly encourage children to pursue excellence, winning, and even to compete with others, this atmosphere may result in that parents become more vulnerable to keep their sensible parenting principles, control behaviours and regulate emotion well. Parents may loss ability to differentiate their own needs for success and achievement from a child’s developmental needs and goals.  They conceptualized the phenomenon of “objectification of child” and “achievement by proxy distortion (ABPD) to emphasize that objectifying children becomes one of parenting approaches to achieve parental “achievement by proxy”. That is, ambitious and competitive parents regarding their child as an object, rather than a person, as a means to satisfy their own need for achievement. With the increasing intensity of parental pressure in sport or other achievement fields, children may cope with this by emotionally distancing themselves from their own needs, colluding in this objectification of themselves. It leads children to feel guilt or lose self-worth if they cannot achieve parents’ high expectations and feel impulsive to succeed to please parents, because they define or value themselves by parents’ recognition and approval of their achievement.

Besides, several researchers also proposed the concept of “parental conditional regards (PCR)” to illustrate a dysfunctional parenting practice which has been prevalent in children’s sport participation and schooling — parents’ love and affection is contingent on children’s specific behaviours that parents desired. This defective parenting way has been abundantly documented its negative correction with children’s self-esteem, personal exploration, self-regulation, emotion control, prosocial behaviour, and academic and sport achievement (Roger, 1951; Miller, 1981; Baldwin, 1994; Deci & Ryan, 1995; Harter, 1993; Assor, Roth, & Deci, 2004). According to previous findings of maladjusted parenting approach especially in children’s sporting and schooling context and its conceptual links with attachment insecurity, we speculate that those unique characteristics of parent-child interactions related to sport and schooling issues in daily life are likely to be typical working models of parent-child “contextual” attachment representations (e.g., sport-specific, schooling-specific, etc.), and those context-specific models become exemplars of a “global” parent-child attachment schema.

Future Directions

• Characterize the activation and operation of the attachment system

1. Children in the context of sport and schooling? other contexts? Possible contexts in adolescence or adulthood?
2. Longitudinal studies needed
3. Measurements for context attachment within a given relationship
4. Can prototype-like context exist?
5. Apply priming exercise in a contextual level (clinic approach)
6. Cultural-different considerations