Introduction
Attachment theory provides a framework for understanding how attachment bonds formed in infancy affect relationships and psychological wellbeing across the lifespan. Indeed, Bowlby claimed that attachments play a ‘vital role’ from the cradle to the grave’. In this paper, I first describe some fundamental ideas pertaining to the theory. I then critique Bowlby’s claim by evaluating literature that pertains to stability of attachment over time. This paper concludes that despite some methodological weaknesses and seemingly inconsistent findings, there is some evidence to suggest that attachment remains relatively consistent throughout the lifespan, but is open to change in certain contexts. I discuss attachment theory in relation to an issue that is more prevalent among older adults; dementia. I then consider the implications for clinical psychology research and practice, by discussing the relevance of attachment theory for working with adults and older adults with mental health difficulties. Throughout this paper I will be interpreting ‘stable way of relating to others’ as ‘stable across time’.
Attachment theory
Bowlby’s attachment theory describes a child’s bond to his/her primary caregiver, and outlines the effect of this bond on subsequent relationships and psychological adjustment throughout the lifespan. Bowlby (1982/1969) proposed the existence of an "attachment system’. This system is designed to promote infants’ survival through maintaining their proximity to their care-giver in times of threat or danger. Bowlby proposed that natural selection would have favoured mammals who formed strong emotional bonds to their care-givers. This would be particularly important in human infants given the extended period of physical immaturity and dependence. Bowlby argued that though the need for the attachment system is obviously at its greatest during infancy, it remains active throughout life, and is evident in the strong feelings, thoughts and behaviours we experience towards loved ones in times of need. In modern life, when we experience hardships, stress or trauma our attachment system is activated and we draw closer those who care for us. Our loved ones provide us with the sense of security that protects our emotional wellbeing and mental and physical health.
As infants our primary care givers are our main attachment figures. However, in adulthood different relationships tend to become more prominent, such as partners, friends, therapists, or God. Attachment theory proposes a hierarchy of attachment figures, with primary attachments being those who we remain intimate with for long periods of time. Adults may not need the physical closeness of an attachment figure in times of threat, as a mental representation of them may suffice (Mikulincer & Shaver 2003).
With this in mind we can understand that Bowlby conceptualised his theory as a way of understanding behaviour throughout the life span, and saw that attachment theory would be relatively stable, but also responsive to subsequent relationships.
Internal working models
Internal working models are important in attachment theory. Bowlby thought that our adult interpersonal styles become embedded through repeated interactions with primary caregivers during childhood. These repeated interactions form cognitive, behavioural and emotional schemas that are designed to maintain proximity to caregivers. As parents have different interpersonal and child rearing styles, so children will have different schemas, or attachment styles, which become ‘roadmaps’ or internal working models for acting in relationships in general (Shorey & Snyder, 2006). For example, if a caregiver is consistent and responsive, a child will learn to trust carers to be supportive when they are in need. These interactions will also shape their understandings of themselves in terms of how worthy they are of care, and therefore how they can expect to be treated by others in the future.
Internal working models are thought to function like ‘self-fulfilling prophecies’, wherein individuals act in ways that precipitate corresponding schema related behaviours from others. These stereotyped interactions then reinforce the validity of the internal working model. Internal working models predispose a person to stable patterns of interactions within relationships. However, they can also produce interpersonal difficulties and mental distress. According to attachment theory, internal working models affect and shape how we see ourselves and how we view and react to others, across the lifespan (Shorey & Snyder, 2006). Indeed, Bowlby claimed, attachments play a ‘vital role’ from the cradle to the grave’ (pg 2 Crowell, Fraley & Shaver, 1999).
A central prediction of attachment theory therefore is that internal working models and attachment schema persist in a relatively stable way over significant portions of the lifetime. However, they are so-called ‘working’ models as they are open to revision in response to significant attachment experiences (Waters, 2000) such as close interpersonal relationships (Rothbard &Shaver 1994 from psychopath paper) or life events that affect care giver responsiveness (Waters, 2000).
Attachment styles
Mary Ainsworth worked to create a classification of internal working models into attachment styles. Mary Ainsworth (1969) created a behavioural experiment to categorise attachment styles, based on infants responses towards their mother. If mothers are emotionally available, relatively consistent and responsive, then infants learn that the mother is dependable and available. They then are then free to explore the world, using the care giver as a secure base, to provide a sense of security and comfort, whenever they feel threatened. These infants are categorised as ‘securely attached’. If, however, the care giver is inconsistent, the infant will be preoccupied with their security needs and hypervigilant to signs of abandonment, thus demonstrating an insecure attachment, specifically the anxious-ambivalent style. If primary care givers are rejecting or absent, a child may learn to supress their emotional need for closeness completely, and an infant will then be categorised as insecure, the anxious-avoidant pattern. Infants displaying disorganised attachment patterns may perceive their caregiver as frightening or hostile, and experience no feelings of being cared for.
In sum, Ainsworth and colleagues created a system to empirically measure infants internal working models, which she called attachment styles. Researchers then were able to test to see if these were stable across time.
Empirical evidence in Infancy and childhood
Several studies using infant samples have confirmed that attachment styles to primary caregivers remain moderately stable across periods of time ranging from 6-7 months in middle class samples (eg Waters 1978, Thompson, Lamb and Estes 1982, Egeland & Sroufe 1981). Research also supports the prediction that changes in attachment from secure to insecure seem to be related to negative life events, where as positive life events are related to changes from insecure to secure (Egeland & Sroufe 1981, Vaughn, Egeland, Sroufe & Waters, 1979).
However, it is important to note that these are only life events that impact on caregiver responsiveness, stressful life events such as moving house don’t seem to influence attachment pattern (Thompson et al 1982).
Moderate stability seems to be less evident in ‘at risk’ samples. Vaughn, Egeland, Sroufe and Water 1979, found that infants who suffered abuse, neglect or maltreatment were categorised more frequently as insecure, and there was less stability in attachment patterns over time than a ‘well cared for’ sample. The authors noted that of the infants who moved from an insecure attachment to a secure attachment, all had some sort of stabilising influence, such as a supportive family member becoming more involved.
To summarise, in line with Bowlby’s prediction, attachment styles to primary caregivers remain fairly stable across 6-7 months of time during infancy, in middle class samples. However, negative attachment related events can cause infants to change attachment style from secure to insecure. An insecure attachment style seems to be maintained by negative attachment events reoccurring, such as abuse, neglect or maltreatment, which are more prevalent in ‘high risk’ samples.
Stability of attachment through infancy to adulthood
Literature looking at stability from infancy to adolescence confirms this view. Two longitudinal studies found significant stability in attachment style, with change from secure to insecure attachment, or maintaining an insecure attachment style, was associated with an attachment related negative life event such as loss of a parent, parent’s mental health difficulty, or abuse by a parent (Hamiliton 2000, Waters 2000).
Weinfield et al 2000 longitudinal study used an ‘at risk’ sample of young mothers, and found significant discontinuity of attachment style, with many participants changing from secure to insecure attachments. The authors found that those that changed from secure to insecure attachments were found to have suffered more attachment related negative life events than the group that were continuous in their attachment style. The sample used were also high poverty and high risk for developmental difficulties, and had experienced a frequency and severity of negative life events than the Hamilton and waters samples discussed above. (Waters 2000).
However, further longitudinal study (Lewis ,Feiring & Rosenthal, 2000), which measured attachment across a span of the first 18 years of life, did not find a high level of consistency in attachment styles, however, individuals from divorced families were more likely to be insecurely attached. This is consistent with the negative attachment related event prediction.
In sum, literature from both infancy and adolescence tell a similar story regarding stability of attachment. Attachment style tends to be relatively stable in middle class samples. However, negative attachment events seem to result in a move from secure to insecure styles, or serve to maintain insecure attachments. Samples that experience more negative life events tend to be more unstable in their attachment styles, as more children move from secure to insecure styles.