Abstract
This study is part of a larger research project that developed the event process model of family violence (FVEPM). The FVEPM was developed by applying grounded theory methods to the event narratives of 14 men and 13 women completing community-based family violence (FV) perpetrator treatment programs. The current study extends this work with the original sample, by examining the routes individual events take through the FVEPM. Three main pathways—comprising 93% of event narratives—were identified: a conflict escalation pathway (n = 14), an automated violence pathway (n = 6), and a compliance pathway (n = 6). Our findings extend existing FV typologies and theories by identifying patterns of features pertaining to the individual, the relationship, and the situation that converge to result in FV perpetration during a FVE. Further validation and development of the pathways may provide FV practitioners with an organizing framework from which to identify more nuanced assessment, treatment planning, and risk management processes for the diverse range of FV perpetrators they are tasked with treating.
The heterogeneity of family violence (FV) perpetrators has long been recognized (Dixon & Browne, 2003; Dixon et al., 2007). In an effort to make sense of this heterogeneity, typologies of intimate partner violence (IPV) have emerged. These typologies largely focus on the individual characteristics and patterns of aggressive behavior of male perpetrators of IPV (Dixon & Browne, 2003). For example, Holtzworth-Munroe and Stuart (1994) developed a typology of male IPV perpetrators based on three dimensions: violence frequency and severity, violence generality, and batterer psychopathology. Family-only batterers perpetrate minor and infrequent physical IPV, only use violence within their intimate relationship, and exhibit little psychopathology. They report limited childhood exposure to FV and partner-specific communication difficulties, do not endorse violence-supportive beliefs, and experience intrapersonal (e.g., emotional regulation) difficulties to a lesser degree than other types. Dysphoric/borderline batterers perpetrate moderate to severe physical, psychological, and sexual IPV, occasionally use violence outside of their intimate relationship, and exhibit psychopathology such as substance abuse, emotional volatility, and personality disorder. They report frequent childhood exposure to FV, violence-supportive beliefs, insecure attachment, and emotional regulation and partner-specific communication difficulties. Finally, generally violent/antisocial batterers perpetrate moderate to severe physical, psychological, and sexual IPV, often use violence outside of their intimate relationship, and exhibit significant psychopathology. They report the highest levels of childhood exposure to FV, a lack of empathy, impulsivity, violence-supportive beliefs, and communication and conflict resolution difficulties across multiple relationships.
Whereas Holtzworth-Munroe and Stuart’s (1994) typology considers IPV at an individual level, Johnson’s (2006) typology takes a dyadic approach. Johnson considers patterns of violence and control by both persons in an intimate relationship: Either one person is violent and controlling (“intimate terrorism”), both persons are violent and controlling (“mutual violent control”), both persons are violent but only one is controlling (“violent resistance”), or one or both persons are violent but neither is controlling (“situational couple violence”). These types align with Holtzworth-Munroe and Stuart’s typology: Situational couple violence resembles family-only batterers, and intimate terrorism resembles dysphoric/borderline and generally violent/antisocial batterers (Johnson, 2006).
By offering insight into different etiologies of IPV perpetration, typologies can provide a framework for identifying more nuanced treatment planning and risk management approaches (Dixon & Graham-Kevan, 2020; Holtzworth-Munroe & Stuart, 1994). Nevertheless, FV researchers warn that typologies in their current form are not well-developed and are too inflexible for their potential clinical utility to be realized (Capaldi & Kim, 2007; Dixon & Browne, 2003). Further, typologies (cf. Johnson, 2006) typically fail to consider the role of relevant situational and interpersonal factors that may usefully differentiate between types (Dixon & Browne, 2003; Holtzworth-Munroe & Meehan, 2004). Indeed, the event process model of family violence (FVEPM; Stairmand et al., 2019) highlights the importance of these factors in contributing to acts of FV.
Among other forensic populations, heterogeneity has been captured by examining distinct pathways in the offending process. Specifically, Ward and colleagues (Murdoch et al., 2012; Polaschek et al., 2001; Ward et al., 1995) have developed models of the offense process that present a temporal outline of a specific offense, including its cognitive, behavioral, contextual, and motivational components. Offense process models are developed by gathering first-hand accounts of the offense process, and then analyzing these accounts using grounded theory methods (Ward et al., 1995). Following model development, potential pathways through the model can be examined. Although offense pathways are descriptive in nature, they have important theoretical and practical implications in that they may help to identify distinct offending etiologies and treatment needs (Polaschek & Hudson, 2004). For example, Polaschek and Hudson’s (2004) offense process model of rapists identified three pathways of the offense process according to offenders’ dominant goals for seeking sexual gratification: to enhance positive mood, to escape negative mood, or to redress harm to the self. These pathways highlight differences in the offense process (e.g., offense planning, denial and cognitive distortions, coping styles) that suggest the presence of distinct treatment targets. Although offense process models have been applied to other forensic populations, they have rarely been applied to FV perpetrators (cf. Drummond, 1999).
The current study is part of a larger research project that developed the FVEPM (Stairmand et al., 2019) by using grounded theory methods to systematically analyze perpetrators’ narratives of FV events (FVEs). This paper further develops the FVEPM by investigating the presence of distinct offending pathways through the model.
Method
Participants
The event narratives of 14 men and 13 women formed the basis of this study.1,2 All participants were completing a community-based FV perpetrator treatment program at the time of their involvement in the research, either on a voluntary (n = 13) or mandated basis through the criminal (n = 9) or family (n = 5) court. Participants identified as New Zealand Māori (n = 12), New Zealand European (n = 8), Pasifika (n = 2), or as having multiple ethnicities (n = 5). Participants ranged in age from 22 to 50 years (M = 34.44, SD = 7.52). Ten participants were unemployed, seven worked full-time, four worked part-time, five were stay-at-home parents, and one studied full-time. Nearly two-thirds of participants reported experiencing physical IPV victimization by the event victim prior to the FVE. 3
Procedure
Participants were recruited from three different FV service providers in the North (n = 2) and South (n = 1) Islands of New Zealand. Participants were recruited either directly through the first author’s (MS’s) attendance at their treatment program (n = 8), or indirectly through being informed of the research by their treatment provider (n = 19). Regardless of the initial recruitment procedure, MS met individually with each potential participant in a private room at the treatment agency. After providing informed consent, participants took part in a semi-structured interview that ranged in length from 46 to 120 minutes (M = 78 minutes). Participants were asked to provide a detailed description of a specific FVE in which they perpetrated FV, as well as any factors they perceived to be important in understanding why the FVE occurred. Other than being asked to describe a FVE they remembered well, participants were given no instructions as to which FVE they should describe. Participants received a $30 voucher for their participation.
Data Analysis
Development of the event process model of family violence
The FVEPM forms the basis of the analysis for this paper; as such, a brief description of its development is provided here. 1 Participants collectively described 32 FVEs in which they used physical (and often also psychological; n = 28) or only psychological (n = 4) FV towards an intimate partner (n = 26), child (n = 3), sibling (n = 2), or parent (n = 1). Although most participants described one FVE, five participants described a second FVE in enough detail for it to be included in data analysis. Preliminary analysis revealed that FVEs involving physical FV and only psychological FV were conceptually similar, as were FVEs that involved partners and non-partners as event victims. As such, all event narratives were included in data analysis.
MS carried out data analysis using NVivo software. Grounded theory methods (Strauss & Corbin, 1990) informed the model development. We selected grounded theory because it privileges a bottom-up approach to theory development, seeks to explain variation in human behavior, acknowledges the importance of context in understanding action and interaction, and is a particularly useful approach in the absence of other theories that adequately explain the phenomena of interest (Strauss & Corbin, 1990). Consistent with a grounded theory approach, interviews were analyzed in sets of one to four to allow for iterative periods of data collection and analysis. After reading each transcript multiple times, MS methodically re-read each transcript line by line to identify, and then label, individual meaning units. New and revised codes were applied to meaning units as data analysis continued, and tentative categories were developed and refined as conceptual links between meaning units began to emerge. Enduring categories were organized into interrelated coding paradigms (i.e., mini-frameworks) that were then grouped into discrete sections based on their temporal relationship to the FVE. Following analysis of the first seven transcripts, a preliminary model of the event process was developed; this preliminary model was revised, tested, and refined during analysis of the remaining 20 transcripts, at which point theoretical saturation occurred.
The FVEPM comprises eight interrelated paradigms within four sections, arranged temporally from the most distal to the most proximal factors in relation to the FVE described (see Table 1). The FVEPM describe participants’ upbringings and early relationship histories (Section 1), their relationships with event victims and the context in which the FVE occurred (Section 2), the FVE itself (Section 3), and the aftermath of the FVE (Section 4). 1
Overview of the Event Process Model of Family Violence (FVEPM).
Pathways analysis
Distinct event pathways that described participants’ progression through each paradigm and section of the FVEPM were identified via a four-stage analysis.
Stage 1. Identifying individual pathways through the model
Using NVivo software, MS prepared a coding report to identify the frequency with which participants endorsed any code within each paradigm and section of the FVEPM at each level of coding (e.g., category, subcategory). Core features of each paradigm were identified, and a coding system was devised to summarize participants’ experiences of these features for each FVE described (see Table 2). Using the coding report and coding system, MS produced a summary table of participants’ progression through the model at the paradigm level (see Table 3). This table formed the basis for the next step of analysis. Four event narratives (FVE 4, 6, 17, and 25; see Table 3) belonging to four participants were excluded from further analysis as there was insufficient information to reliably assign codes across multiple paradigms in Sections 3 and 4. The remaining 28 event narratives belonged to 26 participants. Twenty-four participants described one FVE, and two participants described two FVEs; one FVE involved the same event victim as the first FVE, and one FVE involved a different event victim.
Overview of Pathway Coding System.
Summary Table of Family Violence Events (FVEs).
Note. X = Insufficient information to assign code.
One of two FVEs described by the same participant.
Stage 2. Identifying patterns across participants’ pathways
Next, MS examined each section of the FVEPM in turn to identify shared patterns across participants’ event processes. Within sections, several patterns emerged. For example, participants who reported adverse early events were also likely to report experiencing energized and altered emotions during childhood. Similarly, FVEs characterized by a period of conflict escalation tended not to involve script activation, and vice versa. However, when looking across sections, no meaningful patterns emerged: More than 20 distinct pathways through the model were identified, and 10 distinct pathways were identified for Section 2 alone.
Stage 3. Selecting an organizing framework to guide pathways analysis
Given the heterogeneity observed across individual pathways at the paradigm level, we decided to select a core paradigm or section to function as an organizing framework for further analysis. This strategy was similarly used by Polaschek and Hudson (2004), who organized their pathways analysis according to offenders’ goals. As the FVEPM focuses on one specific FVE (Section 3), and Sections 1, 2, and 4 function to provide the surrounding context for better understanding this FVE, Section 3 was selected as the organizing framework.
Stage 4. Identifying distinct offending pathways through the model
Using the initial coding report generated in NVivo, MS reviewed individual pathways for Section 3 paradigms at the category and subcategory level. Patterns across each individual pathway were then examined. Analysis at this level revealed that the heterogeneity observed within the sample could be usefully captured within the organizing framework of Section 3 paradigms. Three distinct pathways were identified; these pathways accounted for the majority of FVEs, and are described in the results section.
Inter-rater reliability
Following pathways analysis, MS provided coding training to the third author (DP) for the purposes of determining inter-rater reliability. DP independently coded 10 randomly selected event narratives at the subcategory or category (when no subcategories were present) level for Section 3 paradigms. DP then allocated each event narrative to one of the three pathways on this basis. Inter-rater reliability was calculated for each category or subcategory, as well as for pathway allocation. The mean rate of inter-rater agreement was 83.5%. This ranged from 76.5% to 94.1% across individual transcripts, and 60% to 100% across individual categories or subcategories. Categories or subcategories with the least inter-rater agreement were: “Reached a threshold for tolerating the event victim’s behavior,” “Raise and avoid,” “Violence supportive cognitions,” and “Act: Demand.” Mean inter-rater agreement for pathway allocation was 80%. Discussion between MS and DP resulted in all coding discrepancies being resolved.
Results
Of the 28 event narratives for which pathways could be reliably assigned, 26 fit within one of three pathways. One half (n = 14) fit within Pathway 1, and the remaining event narratives were split between Pathway 2 (n = 6) and Pathway 3 (n = 6). The final two narratives involved participants’ use of psychological FV (e.g., damaging the event victim’s personal property) in the absence of any prior interaction with the event victim during the FVE. These narratives shared features of both, but did not fit predominantly within, Pathway 1 or 2. Figure 1 shows patterns in the event process at the paradigm level for each FVEPM pathway. Patterns were considered to occur when at least two-thirds (66%) of event narratives were assigned any given code as per the pathway coding system (see Table 2). As shown in Figure 1, some codes (e.g., “Experienced environmental stressors”) did not appear as a pattern in any pathway, but were endorsed frequently enough by individual participants to warrant their inclusion in the model. Each pathway will now be described, with particular emphasis given to Section 3 paradigms upon which each pathway was based.
Patterns in the event process for each pathway at the paradigm level.
Pathway 1: Conflict Escalation Pathway
Pathway 1 (see Figure 1) comprised 14 event narratives involving 13—six male and seven female—participants and their current or former intimate partner (n = 12) or child (n = 2). Pathway 1 participants described growing up in a violent social environment in which they frequently witnessed physical violence. Participants also directly experienced a range of FV-related adverse early events, including psychological, physical, and sexual FV. Within the context of their failed support seeking, participants were required to independently manage adverse early events and their sequelae. Participants used a range of strategies in an effort to end these events, and to facilitate the short-term and long-term avoidance of the negative emotions that the events evoked. Participants described experiencing parenting practices (e.g., the avoidance or aggressive management of negative emotions and interpersonal stressors) that prevented them from learning how to control and pro-socially express their emotions. Subsequently, some participants developed “Others will hurt me,” “Violence is acceptable,” and “I am worthless” schemas.
Participants’ and event victims’ relationships were characterized by a “honeymoon period,” after which relationship stressors began to arise. The dyadic conflict resolution and communication strategies used to manage these relationship stressors facilitated ongoing relationship conflict. For example, discussion-based strategies involved one or both persons’ attempts to avoid verbal communication about relationship stressors, or immediate escalation into a verbal argument in which both persons refused to consider the other’s perspective and vehemently defended their own. These strategies contributed to a backlog of relationship stressors, characterized by either the reoccurrence of a relationship stressor or the repeated use of ineffective discussion-based strategies to manage existing stressors. Discussion-based strategies could escalate to psychological and—minor or infrequent—physical FV by one or both parties. Enduring efforts to control the other over time—as opposed to temporary and situation-specific efforts during a FVE—were infrequently reported. Although ongoing conflict was a defining feature of these relationships, physical FV was not. In fact, participants reported their deliberate intention to not use physical FV towards the event victim. Some participants experienced multiple and compounding environmental stressors in the lead-up to the FVE.
Pertaining to the FVE itself, participants initiated verbal interaction with the event victim following their evaluation of the event victim’s unacceptable behavior. This evaluation related to a new occurrence of an ongoing relationship stressor, or the event victim’s use of physical or psychological FV: The school rings me up at 4 o’clock, ‘Are you going to pick your children up from school?’ And I went ‘No cause I’m out of town, their father’s meant to pick them up’ . . . I was fuming that he could forget about our children—P19, female
Having recently experienced relationship stressors, environmental stressors, or both, participants described having limited emotional and cognitive resources to manage the impending conflict. Participants’ depleted internal resources influenced their intrapersonal and interpersonal responses during the FVE: [describing suspicions of event victim’s infidelity prior to the FVE] . . . [I was] really upset and, just wondering what to do about it, if it’s true or not or if it’s happening. It was just like a constant battle . . . I felt like I was just losing myself sort of thing. I was losing what sort of dude I am what sort of guy I am—P11, male
Following their initial evaluation, participants selected a strategy for managing their interaction with the event victim. Participants invariably reported an initial intention to ensure the event victim’s compliance, based on their goal of obtaining access to valuable resources and experiences (e.g., information, personal belongings) or ensuring their own and other’s psychological safety or wellbeing. Importantly, participants’ initial attempts to gain compliance involved verbal acts (e.g., contingent threats, verbal demands): I said ‘Stop sending me these f***ing messages, I’ve had enough’ [act: verbal demand] . . . I wanted him to hear me, how it was for me from my point of view and that I need him to stop [intention: compliance] . . . I just wanted to be left alone in peace, I wanted to be happy and I felt at the time that I’m never gonna be happy as long as this guy’s in my life [goal: psychological safety and wellbeing]—P5, female
Event victims responded to these acts in a manner that was incompatible with participants’ goals; for example, by refusing to comply with participants’ demands or using psychological or physical FV towards the participant. Event victims’ responses precipitated participants’ own counter-escalation; for example, by escalating the severity of acts used to achieve their goal or by shifting from a compliance to a harm intention. As such, participants reported that their strategy selection was a dynamic process, characterized by their reselection of multiple and changing acts, intentions, and goals as the conflict escalated: I was trying to talk with [event victim] [act: verbal demand] he just blatantly ignored me . . . I was trying to get him to go outside [intention: compliance] . . . he was being disrespectful and wouldn’t listen so I pulled his ear [act: bodily force]—P12, female
During this process, participants experienced violence-supportive cognitions (e.g., “[Event victim] is hurting me on purpose”) and intensifying anger that they felt increasingly unable to control: I felt like [event victim] was setting me up, like to get rid of me or to, move on with life, and I was, it’s just all this strong emotion of anger . . . I just started seeing red and, just couldn’t I just had no way to control it—P3, male
Eventually, a further perceived transgression—including physical or psychological FV—by the event victim precipitated participants’ physical FV perpetration. Participants’ FV perpetration was accompanied by their newly formed intention to physically harm the event victim, in an effort to achieve retributive justice: [Event victim] pushed me to full blown extent . . . she defended [third party] and that’s when I snapped . . . I just f***in hit her cause that was the only way I could feel that I could hurt her the way that she’d been hurting me—P28, female
Participants perpetrated a range of physically violent acts, from a single push to a sustained physical attack involving multiple blows to the event victim’s head and body. These acts were uncharacteristic—both in terms of presence and severity—of the participant in their relationship with the event victim, and did not result in physical injury. Nevertheless, most FVEs were witnessed by participants’ and event victims’ children.
Participants’ physical FV perpetration signaled the ending or imminent ending of the FVE, either because the participant promptly left the scene or because formal or informal intervention was sought. Formal intervention seeking by the event victim or a third party resulted in the participant receiving legal consequences (e.g., being charged with an offense). Most participants viewed the FVE in a negative light, largely based on the perceived unacceptability of their actions but also based on the non-realization of their goal and the potential consequences of their actions to themselves.
Pathway 2: Automated Violence Pathway
Pathway 2 (see Figure 1) comprised six event narratives involving six—five female and one male—participants and their intimate partner (n = 5) or child (n = 1). Pathway 2 participants’ upbringings were largely indistinguishable from those described in Pathway 1. However, several differences emerged: Pathway 2 participants were more likely to disclose sexual abuse, to report feeling less valued or loved than their siblings, to experience unsuccessful support seeking (e.g., because they were blamed or not believed), and to develop “Others will hurt me” and “Violence is acceptable” schemas.
Pathway 2 participants described a “honeymoon period” and high level of investment (e.g., moving in together, falling pregnant) early in their relationship with the event victim. Most participants experienced severe physical and psychological FV—either as a primary victim (n = 3) or mutual perpetrator (n = 1)—in their relationship with the event victim. Participants’ “I am worthless” schemas, desire to attain an idealized vision of a happy family not present in their own childhood, and event victims’ promises to change and repent facilitated participants’ decisions to endure FV by the event victim. Alternatively, two participants described relationships with event victims characterized by the absence or infrequent use of physical FV. These relationships were also characterized by the absence of meaningful communication; that is, participants and event victims “existed around each other” in a shared space.
As in Pathway 1, FVEs began with the participant’s evaluation of the event victim’s behavior as unacceptable. These evaluations pertained to event victims’ current or recent use of FV, or perceived transgressions by event victims (e.g., confirmed infidelity) that threatened the continuation of the relationship. These transgressions elicited participants’ intensely energized emotions, particularly anger and its physiological components: I walked into [friend’s] room to find her in bed with [event victim] . . . I just felt so enraged like, a rage inside it was painful. It was so painful inside, raw . . . I felt sick in my stomach and I had a burning feeling in my chest and my hands were shaking. Like I was having anxiety. And I could feel my face red, like a fire engine red—P6, female
Participants reported entering the FVE with depleted emotional and cognitive resources; this pertained to their prior experiences of FV victimization. Other participants reported that the lead-up to the FVE was “just an ordinary day”: I knew something was wrong there like I was holding a lot of anger which I am going through with my psychologist, she said it’s built in anger from everything I’ve been through . . . I was just way past my point of, way past my point. I don’t think anything would’ve helped that day. I think it was just a breaking point for me—P7, female
Upon experiencing intensely energized emotions, and in the absence of any attempt to engage in verbal interaction with the event victim, participants perpetrated severe physical FV (e.g., choking, repeated blows to the head and body) that caused physical injury (e.g., loss of consciousness, a severed limb) to the event victim. All but one participant reported their physical harm intention, based on their goal of achieving retributive justice for the physical or emotional hurt the event victim had inflicted on them (“[I] physically smashed them up like threw them to the walls, threw them down the stairs, strangled them both till they turned purple . . . I got hurt so many times so I want them to hurt”—P2, male).
Participants’ accounts of the FVE appeared to be consistent with an automated, script-driven process. Although participants could retrospectively describe their strategy selection, they distinctly recalled the absence of any cognitions or conscious goal formation during the FVE (“There was just no, no thinking . . . there was no thoughts going on in my head . . . there was no goal or anything”—P6, female). Participants’ descriptions of a “blank mind” suggest the unconscious activation of a “Hurt the person who hurt me” script. Due to the unconscious nature of their cognitive processes, participants felt as if they were acting on auto-pilot while perpetrating physical FV (“[Physical FV] was automatic. [My] arms just started swinging straight away . . . My mind just went totally blank. Everything went blank and I was just swinging my arms everywhere”—P7, female). Participants experienced dissociative symptoms during the FVE, including flashbacks to a previous traumatic event, memory blackouts, and depersonalization: I was almost outside of myself actually. It’s like, it was a surreal feeling, very odd . . . I, became very, disconnected from my body pretty much. And, so when I look back now I was looking at everything, but . . . it was like I was looking at, a stranger . . . looking at a stranger who had done this awful thing, to another stranger . . . I was in this other world where, where it wasn’t my life—P21, female
Physical intervention by a third party or a tangible reminder of the current situation (e.g., seeing the event victim’s blood) signaled the end of the participant’s physical FV; this usually coincided with the end of the FVE. Participants left the scene immediately after the FVE. No other patterns emerged regarding the aftermath of the FVE. Participants were equally likely to evaluate the FVE in a positive or negative light, and to receive—or not receive—externally-imposed consequences.
Pathway 3: Compliance Pathway
Pathway 3 (see Figure 1) comprised six event narratives involving five male participants and their current or former intimate partner. Pathway 3 participants’ upbringings were characterized by the absence of (n = 2) or opportunity to escape from (n = 3) exposure to adverse early events and a violent social environment. This escape was either permanent (e.g., a long-term arrangement in a deliberate effort to remove the participant from a violent home), or temporary (e.g., the ability to spend time with loving and prosocial grandparents when “things got tough” at home). As such, participants did not report the range of detrimental Section 1 experiences that characterized Pathways 1 and 2.
Participants’ and event victims’ relationships were characterized by the absence of physical FV and the use of discussion-based strategies that promoted conflict resolution. However, one participant reported their frequent use of physical and psychological FV, and another reported the event victim’s severe but relatively infrequent use of physical FV. Participants reported their dislike of physical FV, including their own use of FV towards the event victim. In the lead-up to three FVEs, participants experienced ongoing relationship stressors or compounding environmental stressors (e.g., multiple deaths in the family) that caused them to experience significant psychological distress.
Pertaining to the FVE itself, participants and event victims were equally likely to initiate verbal interaction based on their perception that the other’s behavior (e.g., substance use) was adversely impacting themselves or their children (“[Event victim] can’t be doing that sort of shit in front of our kids. You know that aint a, you’re setting a bad example”—P15, male). Participants reported their reluctance to engage in prolonged interaction with the event victim; however, they deemed a brief interaction to be necessary in order to achieve a desired outcome. As such, participants’ initial strategy selection involved verbal demands, compliance intentions, and goals of attaining access to valued resources and experiences (e.g., information, solitude) or ensuring their own or others’ safety or wellbeing: [Event victim] wanted me to help out a bit more with the children. And I was, just wanted to sit on my couch and watch movies [goal: access to valuable resources and experiences—solitude] . . . That’s what I said too, “Just leave me alone, get outta my face” [act: demand]. I just, wanna be left alone [intention: compliance]—P1, male
Participants who had recently experienced environmental or relationship stressors reported that their strategy selection occurred within the context of their depleted emotional and cognitive resources (“I was just sort of feeling a bit sorry for myself”—P1, male). As in Pathway 1, event victims responded in a manner that was incompatible with participants’ goals. This resulted in a period of conflict escalation characterized by participants’ attempts to end, and event victims’ persistent efforts to maintain, the current interaction. In four FVEs, event victims initiated physical FV (e.g., pushing, attempting to hit or stab) towards the participant. Participants responded to event victims’ continued demands or physical FV by perpetrating single, minor acts of physical FV (e.g., pushing, physical restraint) towards the event victim; these acts did not cause physical injury. In FVEs involving event victims’ physical FV use, participants’ compliance intentions were now based on their goal of ensuring their own physical safety: I started ignoring [event victim] and she decided to get my attention she was going to run across the room and try and punch me in the head . . . I stood up off the couch, grabbed her, put her out the front door [act: bodily force] . . . It was to stop her [intention: compliance] trying to scratch me . . . I had to manhandle her outside [act: bodily force] kicking and screaming she was trying to scratch my eyes out, kick me in the nuts and bite me [goal: physical safety and wellbeing]— P25, male
Importantly, participants’ physical FV was invariably driven by their unchanging intention to elicit compliance from—not harm—the event victim. All participants reported that their physical FV occurred in the absence of their violence-supportive cognitions and escalating anger (“It was like there was no feeling. . . there was no frustration . . . I just [used physical FV] because I knew that . . . [it] was a form of me getting what I want”—P9, male).
Following the FVE, participants made no considered evaluation of the FVE, or viewed the FVE in a positive light based on the perceived acceptability of their physical FV perpetration within the current context (e.g., the event victim’s initiation of physical FV). No other patterns emerged in the aftermath of the FVE.
Discussion
Analysis of pathways through the FVEPM identified three distinct pathways of FV perpetration. We begin this discussion by considering the key features of each pathway and their relationship to existing psychological and FV research and theories. We then consider potential treatment implications, before discussing the study limitations and avenues for future research.
Pathway 1 (n = 14) was the most common pathway. Pathway 1 participants described childhoods characterized by exposure to FV and dysfunctional parenting practices, and relationships with event victims characterized by ongoing relationship conflict, yet the absence or infrequent use of physical FV. The FVE itself was characterized by an extended period of conflict escalation—sometimes involving the event victim’s use of psychological or physical FV—during which participants experienced escalating anger, violence-supportive cognitions, and a dynamic process of strategy (re)selection. Participants’ initial strategies involved compliance intentions and non-physically violent acts. As the conflict escalated, participants’ newly formed intentions to harm the event victim—based on their revised goal of achieving retributive justice—resulted in their physical FV perpetration. Following the FVE, participants evaluated their actions in a negative light. Participants’ accounts of the FVE and of their relationships with event victims suggest that they perceive themselves as doing their best—but occasionally failing in their endeavor—to not use physical FV. This pathway is consistent with Holtzworth-Munroe and Stuart’s (1994) family-only batterers, whose infrequent IPV perpetration is attributed to their emotion regulation and partner-specific communication difficulties—in combination with their experiencing of personal and relationship stressors—following an extended period of conflict escalation. However, in contrast to Holtzworth-Munroe and Stuart’s type, Pathway 1 also highlights the need to consider the role of dyadic interaction patterns during a FVE.
Pathway 2 participants (n = 6; nearly all women) reported upbringings similar to those described in Pathway 1. In contrast to Pathway 1, Pathway 2 participants tended to describe ongoing experiences of FV victimization in their relationship with the event victim; three participants received, and one participant both perpetrated and received, frequent and severe physical and psychological FV. The FVE was characterized by participants’ intensely energized emotions upon experiencing a perceived transgression by the event victim, prompting their immediate and script-driven use of physical FV with the unchanging intention of delivering physical harm. Participants’ FV perpetration was accompanied by their experiencing dissociative symptoms. Following the FVE, participants were equally likely to evaluate their actions in a positive or negative light. Participants’ accounts of the FVE and of their relationships with event victims are somewhat consistent with Holtzworth-Munroe and Stuart’s (1994) dysphoric/borderline batterer, particularly regarding their high level of investment in the relationship and their script-driven use of physical FV during the FVE to express their intense psychological distress upon experiencing a perceived transgression by the event victim. However, a distinct hallmark of Pathway 2—not well-captured in Holtzworth-Munroe and Stuart’s typology of male perpetrators—is participants’ ongoing experiences of FV victimization. Although two participants did not experience FV victimization in their relationship with the event victim, they continued to relive their prior experiences of FV victimization through intrusive (e.g., flashbacks, nightmares) and other (e.g., avoidance of trauma-related stimuli) symptoms of posttraumatic stress disorder (American Psychiatric Association, 2013). Given their common experiencing of dissociative symptoms during the FVE, and that three FVEs were precipitated by event victims’ FV use, participants’ victimization histories are an integral part of understanding Pathway 2 FVEs. This finding is supported by a growing body of research demonstrating that a minority of FV perpetrators with trauma histories experience dissociation while perpetrating FV (LaMotte & Murphy, 2017; Simoneti et al., 2000; Webermann & Murphy, 2019).
In contrast to Pathways 1 and 2, Pathway 3 participants (n = 5; all men) described childhoods characterized by the absence of, or escape from, exposure to FV and dysfunctional parenting practices. Participants’ and event victims’ relationships were characterized by the absence of physical FV and the use of discussion-based strategies that promoted conflict resolution. FVEs were characterized by a period of conflict escalation in which participants attempted to end or avoid, and event victims attempted to maintain, the current interaction. In two-thirds of FVEs, participants’ physical FV was preceded by event victims’ own physical FV use. In all FVEs, participants reported their unchanging intention to ensure event victims’ compliance (e.g., to prevent the event victim from inflicting physical harm). Pathway 3 FVEs were distinct from other pathways in several important ways: Participants consistently reported the absence of their escalating anger, violence-supportive cognitions, or a desire to harm the event victim in any way. Accordingly, participants evaluated their physical FV perpetration in a neutral or positive light. For these reasons, Pathway 3 did not map onto any of Holtzworth-Munroe and Stuart’s (1994) types.
The key features of the FVEPM pathways are largely consistent with the results of a previous thematic analysis of victims’ and perpetrators’ offence narratives in a community sample (Testa et al., 2020). In Testa et al.’s (2020) thematic analysis, the majority of offence narratives were characterized by minor physical IPV (e.g., a single slap or push). All but one offence narrative could be classified according to one of three primary functions of physical IPV: Expressive, Punishment, and Instrumental. Consistent with Pathway 1, Expressive narratives involved the use of physical IPV following an extended period of conflict escalation in which participants reported their increasing anger, loss of control, and perceived inability to express their intensifying emotions using verbal means. Consistent with Pathway 2, Punishment narratives involved the immediate perpetration of physical IPV in an effort to punish a partner following a perceived serious transgression. However, the authors did not describe participants’ script activation and dissociative experiences that were a key feature of Pathway 2. Consistent with Pathway 3, Instrumental narratives involved the use of physical IPV to achieve a specific purpose (e.g., to make a partner leave) in the absence of escalating anger and conflict escalation.
The general strain theory (GST) of crime and delinquency (Agnew, 1992) provides further insight into patterns of FV reported in our research. Specifically, the GST argues that experiencing strain will likely elicit negative emotions—particularly anger—within an individual, thereby pressuring them to commit delinquent acts. Strain is conceptualized as any action by others that (1) stops the individual from achieving a valued goal, (2) results in the individual losing a positively valued stimulus, and (3) results in the individual experiencing an unwanted stimulus. The GST recognizes that delinquent behavior is just one potential method of alleviating strain; the likelihood that this behavior will be carried out depends on the nature (e.g., chronicity, magnitude, cumulation) of the strain, as well as features pertaining to the individual themselves. These features include constraints on delinquent and other forms of coping (e.g., goals and values, coping resources, social support), and dispositions to engage—or not—in delinquent behavior (e.g., temperament, learning history, schemas, and attributions relating to the strain).
Key features of GST are evident in each of our pathways, but are particularly salient in Pathway 1 in which participants entered the FVE with depleted emotional and cognitive resources, having experienced chronic and cumulating strains (e.g., a backlog of relationship stressors) and initial unsuccessful attempts to cope with this strain (e.g., failed support seeking, individual coping strategies to manage relationship stressors) in the lead-up to the FVE. Although Pathway 3 participants also experienced these strains, their relatively prosocial dispositions and lack of constraints towards alternative forms of coping perhaps prevented these FVEs from more closely resembling a Pathway 1 FVE. Further, Pathway 2 FVEs highlight the need to understand the subjective nature of the strain to the individual; the perceived magnitude of the strain initiated an automated process, resulting in participants’ perceived inability to engage in anything other than physical FV.
The GST is not dissimilar from theories of FV in that it identifies that delinquent—or FV—acts are carried out by some individuals some of the time (Agnew, 1992). In particular, the ecological-transactional model of child maltreatment (Cicchetti & Lynch, 1993) and I3 model (Finkel, 2008) argue that multiple vulnerability and protective factors—both enduring and situation-specific—contribute to FV perpetration. When the strength of vulnerability factors is cumulatively greater than the strength of protective factors, FV perpetration is likely to occur. These two FV theories share a key limitation of the GST: They have difficulty explaining which individuals, in which contexts, will perpetrate FV (or in the case of the GST, delinquent acts; Agnew, 2013). An event-based approach such as this enables the presence, interaction, and temporal relationships between protective and vulnerability factors to be identified, both during and in the lead-up to a FVE. As such, our research extends existing FV theories by identifying patterns of features pertaining to the individual, the relationship, and the situation that likely converge to result in FV perpetration.
Turning now to potential treatment implications, our pathways analysis suggests that participants’ treatment needs are broadly consistent with those already targeted in cognitive behavioral FV treatment programs. Across pathways—albeit to a much lesser extent in Pathway 3—participants’ event narratives suggest that they would benefit from developing and practicing emotion regulation skills, communication and conflict resolution skills, and effective coping strategies for managing interpersonal and life stressors. Given participants’ repeated experiences of unsuccessful and untried support seeking, participants could also be assisted by treatment providers to gain access to, and increase their willingness and ability to use, formal and informal support networks. Across all three pathways, participants reported experiencing physical IPV victimization by the event victim both during, and prior to, the FVE. Prior to the FVE, participants reported experiencing minor or infrequent physical and psychological FV victimization in Pathway 1, severe and frequent physical and psychological FV victimization in Pathway 2, and no physical or psychological FV victimization in Pathway 3. Given the association between IPV perpetration and victimization (Stith et al., 2004), and the high rates of bidirectional IPV reported in community treatment samples (Wray et al., 2016), participants’ own risk of FV victimization should arguably be assessed. Similarly, participants’ accounts of FVEs suggest that where appropriate and safe to do so, some event victims would also benefit from gaining access to intervention programs in which they could learn and practice the same skills.
Specific to Pathways 1 and 3, FVEs were characterized by an extended period of conflict escalation preceding participants’ physical FV perpetration. This highlights the multiple potential opportunities for de-escalation during a FVE. Along these lines, treatment programs should assist Pathway 1 and 3 participants to identify and manage high-risk situations in an effort to de-escalate interpersonal conflict and avoid physical FV perpetration. Specific to Pathway 2, participants’ ongoing experiences of FV victimization suggest that their engagement in FV treatment programs should be trauma-informed. Here, treatment providers should acknowledge and provide participants with the opportunity to process their experiences of FV victimization (e.g., through referral to specialist treatment providers) before—or alongside—engaging in treatment focused on their FV perpetration. Brief periods of effective treatment (e.g., EMDR-based treatment; Stowasser, 2007) may enable these participants to gain better control over their responses to triggers in their current relationships, thereby enabling them to make better use of other strategies regularly taught in FV perpetrator treatment programs. Given Pathway 1 participants’ experiences of childhood FV victimization, they too may benefit from a trauma-informed treatment approach.
Pathway 1 and 3 participants described their attempts to ensure event victims’ compliance during the FVE. Power and control motives are routinely argued to provide support for Duluth-based perspectives, in which all forms of FV are viewed as a deliberate, gender-specific tactic used by men to oppress and control their female partners. However, our analyses suggested that participants’ efforts to ensure event victims’ compliance during a FVE were driven by a range of factors, including access to valued resources and physical or psychological safety. As such, and in contrast to Duluth-based perspectives, participants’ efforts to control event victims’ during a FVE better represent a one-sided conflict resolution tactic to elicit situation-specific compliance.
Interestingly, no pattern emerged regarding participants’ more general desire to exert power and control over event victims. Across pathways, a small number of participants described their more enduring attempts to control event victims (e.g., through social isolation, monitoring behavior, and restricting financial independence, personal freedom and decision making) prior to the FVE. However, participants were as likely to attribute these more enduring attempts to control event victims as a—misguided—coping strategy to manage their personal (e.g., low self-esteem) and relationship (e.g., a fear of being cheated on) insecurities as they were to their socially-constructed beliefs regarding the inferiority of women. This finding supports the need to consider a variety of explanations other than patriarchal beliefs for patterns of controlling behavior in intimate and family relationships. For example, acts of FV driven by participants’ jealousy and fears of abandonment may better represent attachment anxiety than gendered beliefs (Barbaro et al., 2019; Cheche Hoover & Jackson, 2019).
There are several limitations to this study. First, our pathways are constructed from a small sample of participants who reported that they perpetrated isolated or infrequent acts of physical IPV. Participants who characteristically used physical FV towards event victims were not well represented in our sample; it is therefore likely that other pathways to FV are not accounted for by this sample. Although this study did not set out to provide a comprehensive account of all FV pathways, theoretical and empirical accounts of FV—including typology research (Holtzworth-Munroe et al., 1994; Johnson, 2006)—highlight the importance of better understanding this latter group of perpetrators. Investigation of this group is therefore warranted in further research.
Second, our analysis does not escape the limitations of typology research in that event pathways are primarily based on participants’ behavior at one point in time (Capaldi & Kim, 2007). The stability of these pathways across an individual’s FVEs, and over the course of the relationship, remains unclear. For example, the majority of participants reported describing FVEs involving their most severe use of FV within the history of the relationship; a typical FVE may have fit within an entirely different pathway. Relatedly, the decision to use Section 3 as an organizing framework for pathways analysis limited our ability to more clearly examine patterns in participants’ and event victims’ relationship history. These patterns are well-captured by Johnson’s (2006) typology. For example, Pathway 2 participants described relationships that could be classified as mutual violent control, violent resistance, and situational couple violence according to Johnson’s (2006) typology, as well as those that were characterized by the absence of physical FV.
Third, there are limitations associated with participants’ self-reports, including the possibility that they may be minimizing their FV perpetration, and will likely view the FVE differently to the event victim (Heckert & Gondolf, 2000). Nevertheless, participants appeared willing to disclose information that may portray them in a negative light, including information that they had not previously disclosed to police or treatment providers. Participants were also able to provide detailed accounts of intrapersonal processes (e.g., dissociative states, violence-supportive cognitions) not accessible to others. FV research rarely seeks to obtain perpetrators’ perspectives, despite the obvious theoretical and clinical utility of understanding how perpetrators make sense of their own behavior (James et al., 2002). In many ways, then, relying on participants’ subjective accounts also represents an important strength of our research.
Potential avenues for future research include further pathway development and validation by expanding our sample to include more perpetrators of non-IPV forms of FV, as well a high-risk group (extension currently underway). Given that the majority of participants described only one FVE, future research could also determine the stability of pathways across multiple FVEs. Relatedly, researchers could examine potential differences between pathways in event characteristics over time, such as patterns of desistence or re-offense. Acknowledging its limitations, this study nevertheless offers a unique approach to capturing heterogeneity in the event processes of FV perpetrators. Further, it addresses a major limitation of FV typologies in that it considers the important role of dynamic interpersonal and situational factors in contributing to FV perpetration. These factors are particularly evident in Pathways 1 and 3, in which participants’ physical FV was preceded by event victims’ psychological and physical FV. Further validation and development of the pathways may suggest potential avenues for future research, inform theory development, and provide FV practitioners with an organizing framework from which to identify more nuanced assessment, treatment planning, and risk management processes for the diverse range of FV perpetrators they are tasked with treating.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
