Definitions of domestic violence (DV) have varied over time. However, it is currently recognised that DV is the recurring, random and regular use of abuse to control a partner within an intimate or family relationship (Refuge). This abuse includes physical and sexual assaults, denial, minimisation and victim blaming, threats and intimidation, emotional abuse, humiliation and degradation, use of male privilege, economic abuse, isolation, deception, deprivation, sexual objectification, stalking, using children, and coercion.
At least 1 in 4 women experience DV in their lifetime, whilst between 1 in 8 and 1 in 10 women experience it on a yearly basis. Despite UK police receiving reports of less than 50% of incidents, one DV call is received every minute. In this country, between 16% and 25% of all recorded violent offences involves DV (Dodd, Nicholas, Povey & Walker, 2004). According to Hesse (1999), women are under-represented among victims of violence.
According to Bowlby (1988), attachment is a social bond between individuals, and it includes the behaviours that mediate the development and preservation of that bond. Early attachment with a specific caregiver, usually the mother of the child, ensures he or she is protected and increases the chance of survival. Bowlby (1973) also proposed that variations in the attachment styles of children originate from ‘internal working models’ of the self and other which are shaped from interactions between the infant and the parent. These internal working models can be categorised into two dimensions. The ‘self’ model is distinguished by the extent of emotional dependence on others for self-validation, whilst the ‘other’ model is distinguished by expectations around the availability of others.
Bartholomew (1990) and Bartholomew & Horowitz (1991) developed a four-category model of attachment, which is now widely used in the research of DV: secure – at ease with independence and closeness; fearful – socially avoidant; dismissing: rebuffing attachment; and preoccupied – overly dependent and ambivalent. Fearful and dismissing styles are typified by high attachment avoidance, or discomfort with intimacy associated with a negative other model. Preoccupied and fearful styles are typified by high attachment anxiety, or a fear of abandonment and rejection associated with a negative self model (Corcoran & Mallinckrodt, 2000).
How does Attachment Style influence DV
The most pertinent facet of attachment theory to DV is the concept that anger follows unattained attachment needs. Bartholomew (1990) propose that those with a secure attachment style would not be abusive in intimate relationships as each partner has both a positive self model and positive other model. Consequently, they are both confident and at ease with closeness in their intimate relationships. Those with a dismissing attachment style have a negative other model and a positive self model. They sustain this by defensively minimising the significance of attachment needs and emotionally keeping their distance in relationships. The frustration centred on attachment needs related to a dismissing style may induce anger. However, such people do not tend to experience insecurity in intimate relationships, and therefore, should not be likely to demonstrate anger towards a partner.
Those with preoccupied and fearful attachment styles, defined in terms of a negative self model, are constantly fearful of being deserted or rebuffed in their intimate relationships. Consequently, such anxious attachment styles are predicted to be correlated with high levels of negative influence, including anger and abandonment anxiety in their close relationships (Dutton, 1995).
How has AS previously been measured in relation to DV
Using the Relationship Questionnaire (RQ) to assess adult attachment, and the Conflict Tactic Scales (CTS) to assess the frequency of physical violence, Doumas, Pearson, Elgin & McKinley (2008) observed the association between adult attachment and DV in a sample of 70 couples. Results signified that the combination of an avoidant male with an anxious female was linked to both male and female violence. Whilst the research findings increase the information available in the area, the study is not without its limitations. Firstly, the RQ does not measure anxiety or avoidance dimensions directly; therefore, Doumas et al (2008) suggest that future research should use more recent multi-item measures to overcome this issue. Additionally, the CTS is limited in that it does not measure the severity of violence nor injuries resulting from it, or the issue of violence being used as a form of protection.
Also in 2008, Allison, Bartholomew, Mayseless & Dutton investigated attachment and relationship dynamics in couples were the male has been recognised as a perpetrator of intimate partner violence. Based on the History of Attachments Interview, individuals were assessed for attachment style. Consequently, two attachment strategies for monitoring distance within these relationships were identified – pursuit and distancing. Allison et al (2008) claimed that acts of abuse towards an intimate partner were associated with either one or the other of these strategies. However, in this study there were violent occurrences that did not fit into these themes nor were associated with dynamics between the couples. Furthermore, other themes and dynamics may have become apparent if the population had also consisted of female perpetrators of DV or mutual DV.
The revised Conflict Tactic Scales (CTS-2) and Experiences in Close Relationships Revised questionnaire were used to investigate the severity, reciprocity, frequency and consequences of DV by female perpetrators in Orcutt, Garcia & Pickett’s (2005) study. They discovered that the highest levels of attachment anxiety were found in violent females. Additionally, those with a high anxiety and low avoidance style were more likely to be perpetrators of violence than those high in both attachment styles. The CTS was again criticised for not considering the purpose or wider situation in which aggression transpires. Furthermore, the participants were not specifically asked to consider only one relationship when completing the CTS, and therefore, may have taken into account violence within a number of relationships. Therefore, they may have been the victim with one partner and the perpetrator with another, as opposed to both in the same relationship.
Holtzworth-Munroe, Stuart & Hutchinson (1997) discovered that husbands who were violent towards their wives demonstrated high levels of attachment anxiety, in particular preoccupied and fearful styles, on self-report measures of attachment, and varied insecure styles on the Adult Attachment Interview. Babcock, Jacobson, Gottman & Yerrington (2000), in a replication and extension of the Holtzworth-Munroe et al. (1997) study, stated that husbands who were violent towards their wives were more prone to insecurity than were nonviolent husbands.
According to Gormley & Lopez (2003) as cited in Gormley (2005), an avoidant style of attachment was a better predictor of men’s psychological abuse, whilst Roberts & Noller (1998) claimed that physical violence was perpetrated more by those with an anxious attachment style. Furthermore, Dutton, Saunders, Starzomski & Bartholomew (1994) found a clear association between attachment anxiety and physical abuse, and between both attachment anxiety and avoidance and psychological abuse.
What factors other than AS may influence DV
Dutton’s (1995) nested ecological approach explains domestic violence as multi-determined, taking into account the interpersonal context and the intrapsychic characteristics of the abuser. The model is composed of four levels: the macrosystem, beliefs which stem from one’s culture, such as ‘a women’s place is in the home’; the exosystem, surrounding social structures, such as the influence of peers; the microsystem, immediate environment, such as the level of conflict with an intimate partner; and the ontogenetic level, personal traits, such as learned habits. According to psychodynamic theory, some people are prone to violence due to an ordeal from their childhood or a personality disorder. It claims that the issue of domestic violence should therefore be addressed by identifying the subconscious problem, then consciously remedying it (Healey, Smith & O’Sullivan, 1998).
Family systems theory argues that behavioural problems derive from a dysfunctional family, where everybody is responsible for conflict. Consequently, domestic violence is a result of intimate partners interacting together; therefore, neither is classed as the abuser, regardless of whether only one of them displays physical violence (Healey, Smith & O’Sullivan, 1998). Disempowerment theory states that men who perceive themselves as being inadequate or have low self-esteem are at risk of using violence to assert power (Archer, 1994). Consequently, they redress this by attempting to control those they are threatened by (Gondolf, Fisher, Fisher, & McPherson, 1988).
According to Pro-Feminist theory, male to female domestic violence is due to the patriarchal organisation of society where the male is viewed as the head of the family and other social institutions. As a result of this, perpetrators often believe that they are permitted to control the actions of their partner. Should their authority be threatened, they respond with violence in order to regain control, whilst holding the belief that their behaviour is warranted (Healey, Smith & O’Sullivan, 1998). Cognitive-behavioural theory proposes that domestic violence is due to the re-enactment of abuse from childhood experiences or media influence. It further claims that through abuse, the perpetrator succeeds in obtaining what he wants, which is strengthened through acquiescence on the part of the victim (Hamill, Hayward, Wynn & Craven, 1997).
Rationale for why I chose to research AS
This study is limited to focus on male perpetrators as the effects of anxious or avoidant attachment styles may differ for males and females. Furthermore, there is no research in the area of attachment style and DV with a prison population as past studies were typically community-based. Therefore, I have replicated existing research as previous findings may not be true of the prison population. Finally, as the majority of research in this area is North American, I feel that there is a need for more British based work.
The purpose of the present study was to test a number of hypotheses, the first being that there will be a correlation between an anxious or avoidant attachment style and DV. In addition, those with an anxious attachment style will be more physically violent towards their female intimate partners, whilst those with an avoidant attachment style will be more psychologically violent towards their female intimate partners.
Violence has been defined as both physical and psychological. Physical violence includes any form of unwanted contact between the victim and the perpetrator. For example, hitting, choking, restraining, forcing himself on her.