Abstract
This article presents a categorisation of gaslighting behaviours found in survivor reports of interactions with UK law enforcement in relation to cases involving domestic abuse. This research shows how police interactions can both replicate perpetrator gaslighting behaviours and, consequently, reinforce perpetrator abuse for victims and survivors. A qualitative thematic analysis was applied to a survey of survivor experiences (
Introduction
In this article we examine responses to a survey completed by victims and survivors of domestic abuse across the UK which focused on engagement with police as part of a domestic abuse case. Based on our analysis of these responses, we found that linguistic and operational strategies that some police officers use appear to gaslight victims and survivors. When these strategies are employed, particularly at the point of initial interaction, they can deter victims and survivors from seeking help and, ultimately, justice. We go further in this paper and identify systemic institutional gaslighting strategies that emerged in the responses to our survey. We consider the impact of police gaslighting and argue that when police gaslight victims and survivors, they replicate the behaviours of domestic abusers themselves, resulting in a form of institutional betrayal (Smith and Freyd, 2013) that constitutes a “second assault” (Williams and Holmes, 1981; see also Martin and Powell, 1994) on victims and survivors.
Domestic abuse (DA) is a significant crime throughout the UK. Police data reflects a gendered trend: in the year ending March 2024, females were victims in 72.5% of domestic abuse-related crimes (Office for National Statistics, 2024). Police statistics show that approximately 1.4 million domestic abuse-related incidents and crimes in England and Wales were reported in the year ending March 2024 (Office for National Statistics, 2024). In terms of prosecutions, records show there were 51,183 domestic abuse-related prosecutions in the year ending March 2024 with a conviction rate of 75.8% (Office for National Statistics, 2024). Taken together, the prevalence of domestic abuse and the corresponding statistics show a worrying picture – domestic abuse is endemic, is perpetrated on almost three times as many women as men, is under-reported to police, and of the total cases brought to police, relatively few (3.6%) actually progress through the justice system to prosecution (never mind conviction).
In an effort to tackle the prevalence of domestic abuse and to raise awareness of the complex dynamics of coercive and controlling behaviour (a criminal offence in the UK since 29 December 2015), some strategic interventions have been initiated over the last 18 months. Charitable initiatives include a £2 million funding grant to Women’s Aid, a national DA charity, “to provide one-off payments to survivors of domestic abuse to help them flee, and stay fled, from abusers” (Women’s Aid, 2024). Community initiatives also raise awareness by helping identify and protect community members at risk through a toolkit launched by Neighbourhood Watch (NW): https://www.ourwatch.org.uk/crime-prevention/crime-types/domestic-abuse/domestic-abuse-campaign-toolkit. The creative industries also contribute vital resources; in 2023, actor and DA activist Samantha Beckinsale wrote and produced (with Jason Figgis) a powerful film (“Love?”) that foregrounds the impact of and complex nature of coercive control. These are just some examples of a huge body of work being done to foreground domestic abuse and its physical, emotional, familial, and economic impacts. Despite such valuable learning, the fact is that a disproportionate number of cases are failing to progress through the criminal justice system.
One key area identified for improvement is the way police officers respond to and discursively manage reported cases of domestic abuse (Canning, 2023; Lynn and Canning, 2021). Research has shown that how police officers initially respond to “incidents” of DA can shape whether or not victims feel able to support prosecutions (Barrow Grint, 2016). The impact of initial responses to DA has been outlined in a Government policy paper (Home Office, 2025) which noted that “a positive first response can serve a number of critical purposes – making the difference between a survivor supporting a criminal justice process or not”. What is said and written in the initial stages of DA investigations, therefore, can influence whether or not a case progresses to prosecution. It also influences the building of trust and confidence in policing and in the criminal justice system.
Literature review
“Gaslighting” is a term that is used to describe a form of mental and emotional manipulation by the gaslighter designed to incite the gaslightee to question their perception of reality. It is coercive and abusive. The term “gaslight” derives from a 1938 theatre play of the same name by Patrick Hamilton, later adapted into a film by George Cukor (1944). The plot focuses on a husband’s desire for his wife’s wealth and his subsequent intention to gain it by manipulating her to the point whereby she questions her own sanity. His ultimate aim is to have her certified insane so that he can gain control of her finances and thus, her fortune. The husband manipulates his wife by covertly altering the brightness of the gas lamps in the loft of the marital home and then denying that they are altered when his wife raises this anomaly with him. She then is forced to accept that she is mistaken. Because of her husband’s insistence that the lights are not dimming, she sees no other logical explanation for her apparently warped perception. The term “gaslight”, then, has risen to popularity by operating as a metonym for a more general insidious practice, namely the intentional mental and emotional manipulation in which the gaslighter causes the gaslightee to question their own experiences (cf. Abramson, 2014; Sweet, 2019).
Gaslighting is predominantly enacted through language. In her 2014 paper, Abramson makes this point, albeit obliquely, noting that there are “characteristic things that gaslighters say”. Abramson (2014: 2) makes special mention of “the structure of interactions” that constitute gaslighting. It is worth setting out her points in full: First, the target is framed in the mind of the gaslighter as crazy, paranoid, overreacting or oversensitive—framed in such a way, that is, that she cannot be the source of genuine disagreement. Then the gaslighter tells her this is how he sees her, in the form of a proclamation—e.g. “that’s crazy”—or a command—e.g. “don’t be paranoid”. Then s/he insists on the dismissive framing in their interactions; re-entrenching or using other terms of dismissal if she resists. And finally (though often this is going on throughout), the gaslighter manipulates his target (2014: 14).
While Abramson does not offer a taxonomy of gaslighting, she does provide “illustrat[ions]” (2014: 3) such as those noted above (“that’s crazy”, “don’t be paranoid”). Abramson’s fundamental point is that gaslighting has a characteristic aim – it’s “interpersonal” (2014: 10). Specifically, a gaslighter gaslights, in Abramson’s terms, because of “an interpersonal need for assent, intolerance for challenge or even the possibility of being challenged, and the manipulative destruction of the gaslightee’s standing to issue challenge” (2014: 12). As such, it is perhaps no surprise that gaslighting is a common tactic used by perpetrators of domestic abuse to control and manipulate their victims. Indeed, in Abramson’s study “manipulative leverage” is a “prototypical” feature of gaslighting (2014: 14). Therefore, it is worth analysing what kinds of utterances and pragmatic behaviours constitute gaslighting by police responding to and investigating domestic abuse.
Despite the pervasiveness of gaslighting, its linguistic manifestations are under-researched. Further, gaslighting as a practice within a law enforcement context in relation to domestic abuse cases is an area that has received little to no attention. While police practice in relation to domestic abuse has been researched (e.g. Fleming and Franklin, 2021; Johnson, 2007, inter alia), we are the first to offer direct categorisation of gaslighting behaviours within policing. Darke et al. (2025) provide one of the single most comprehensive overviews of gaslighting literature to date across multiple disciplines, incorporating sociology, medicine, and psychology. Within this, the different foci in the use of “gaslighting” (ranging from interpersonal violence to communication structures) demonstrates holistically the “inconsistent”, varied, and “unclear” application of the term (Darke et al., 2025). Foregrounding these inconsistencies is important, and our paper puts forward an empirically-based operationalisation of gaslighting behaviours, which builds on domestic abuse and intimate partner violence (IPV) research (Bhatti et al., 2023; Darke et al., 2025; Stark, 2019; Sweet, 2019). Darke et al.’s (2025) summary outlines key behaviours and tactics from across the literature, including “manipulation of reality”; “denial”; “inconsistency of behaviour”; “isolation”; “coercion”; and “resulting self-doubt” (Darke et al., 2025, passim). We propose a consistent application of these gaslighting behaviours.
Stark (2019) makes a distinction between “epistemic gaslighting” and “manipulative gaslighting”, which is namely one of “intent”. Stark argues that intent is a determinant of manipulation and disambiguates gaslighting from “reasonable argument” (2019: 224). For an interaction to be considered gaslighting rather than reasonable argument, Stark proffers that gaslighting shows a lack of responsiveness to evidence, instead seeking to “intentionally and methodically (though not necessarily consciously) circumvent” evidence that displaces the gaslightee’s argument and/or supports that of the gaslighter. Sweet (2019: 852) describes gaslighting as “a set of attempts to create a ‘surreal’ (Ferraro, 2006) social environment by making the other in an intimate relationship seem or feel ‘crazy’” and provides a compelling basis for gaslighting to be viewed as sociological rather than just psychological. Sweet’s (2019: 869) view of gaslighting as the enactment of social behaviours (such as “inequalities”, “stereotypes” (passim) and “power relationships”), demonstrates the importance of understanding gaslighting within the context in which it is taking place. Sweet also highlights that although DA/IPV research may include evidence of gaslighting as a “common” behaviour, the term “gaslighting” is not always utilised. To further consolidate this, the findings discussed in this paper conceptually recognise the distinction between “reasonable argument” and “gaslighting” (Stark, 2019: 224). As Sweet (2019: 852) argues, gaslighting is a “social phenomenon” that is context-dependent.
The manipulation of “power” in gaslighting is foregrounded by both Sweet (2019) and Abramson (2014). The aim of gaslighting is to position the gaslightee so that their reality is so untenable that they can no longer rely upon it. Power is an inherent component of coercion (see Abramson, 2014: 19) and that power can be institutional. As will be shown in the findings and discussion below, gaslighting in the context of domestic abuse can extend beyond the domestic perpetrator.
Methodology
The approach we take in our study is qualitative. We developed and shared (through X, formerly Twitter) an online survey inviting statistical and narrative responses to 21 questions that asked about participants’ experience(s) of engaging with police as a result of domestic abuse using our network of domestic abuse charities. We received 235 statistical responses from participants relating accounts of interactions with police and criminal justice services in the UK in addition to extensive, frank testimony from 231 respondents on the physical, emotional, and economic trauma resulting from domestic abuse. Respondents were comprehensive in detailing the processes and impact of escalating cases of domestic abuse through the criminal justice system. Of the 235 respondents, all of whom replied as victims or survivors of domestic abuse, 233 were female. Our interpretive qualitative approach enabled us to analyse all 231 free-text responses using Thematic Analysis (TA) (Braun and Clarke, 2006).
Thematic Analysis (Braun and Clarke, 2006) is an approach that can be applied to qualitative data to help determine the presence and prevalence of thematic patterns. The process includes:
- coding (labelling) the data;
- refining codes;
- identifying code clusters/potential themes;
- refining themes and subthemes based on coded evidence;
- analysing the themes/subthemes to draw conclusions about the data.
The data used in this paper went through a two-cycle coding process split between both authors. In the first cycle, each of us identified codes independently (blind coding), before reviewing and refining them collaboratively. A focused coding approach (Saldaña, 2021) was taken for the second cycle, where categorisation was further refined. These codes were then analysed using Braun and Clarke’s (2006) approach to Thematic Analysis, with data themed categorically (Saldaña, 2021). This allowed us to take an inductive approach to the data, in which no predetermined schemata were applied. Codes and categories were the result of patterns observed within the data. Qualitative approaches are open to criticism, specifically around issues of subconscious bias and subjectivity; however, the rigour with which the process was applied in our study – two-cycle, blind, inductive coding – mitigated this risk considerably.
NVivo 12 was used to code, collate, and categorise qualitative responses across all participants. This allowed researchers to review code clusters holistically and enabled the creation of parent/child “nodes”, which allowed for further categorisation and relationships when establishing connections within the data. Nodes in this instance are the NVivo representations of codes. Parent/child nodes are the representation of code clusters within this programme. The approach of qualitative Thematic Analysis also resonates with the conceptual approach found in critical victimology (Mawby and Walklate, 1994; Walklate et al., 2018) and the emergent narrative victimology (Pemberton et al., 2019b) which give increasing primacy to the experience of victims as told by victims. The voice of the victim has historically been absent in criminology and crime-related research. Our study addresses this lack and develops the narrative turn by foregrounding victims’ and survivors’ own experiences of institutional gaslighting in their own words, keeping them central within victimology research (Pemberton et al., 2019a). Our focus bridges the gap between solely positivist victimology approaches (favouring “aggregate data”) and critical victimology, which seeks to include the victim’s voice (favouring “individual experience”) (Walklate et al., 2018). This interdisciplinary approach adds depth and cohesion to the theoretical underpinnings of such research.
The importance of a victim-survivor voice cannot be overstated. Often in traumatic, isolating, and potentially life-threatening situations, the telling of one’s account can be considered both a “means to regain agency and re-establish communion with their social context” (Pemberton et al., 2019a: 415) and a profoundly brave and enabling act. As will be shown below, victims and survivors of domestic abuse often have their accounts challenged, and consequently are re-victimised (Pemberton et al., 2019a). It is vital, then, for victims and survivors to have a safe space in which to share their experiences in their own words.
In analysing the instances of gaslighting within victims’ and survivors’ accounts, we make recourse to the pragmatic framework of implicature (Canning and Walker, 2024; Clark, 2021; Grice, 1975). Pragmatic implicatures occur when what is said is not a reliable indicator of what is meant. Additional meaning – or a different meaning – is achieved by “blatantly” flouting what Grice has called “maxims” of conversation theoretically required to ensure that interlocutors perceive the intended meanings of utterances in the context (i.e. be relevant, be truthful, don’t say too much or too little, and don’t be ambiguous). As Canning and Walker (2024) put it: When flouting, the speaker wants the hearer to notice that a maxim has not been observed and intends for the hearer to derive an alternative meaning (an implicature) from this flout. This alternative meaning is not explicitly encoded in the language uttered but is implicated by the flout. In other words, the implicated meaning cannot be traced to the form of the utterance. Also, because there is an expectation that the speaker is observing the cooperative principle, the hearer assumes that the speaker’s flout is meaningful and not simply a ploy to derail the conversation just to be mischievous (pp. 142–143).
The implicatures in the gaslighting examples throughout this article carry meanings beyond their conventional semantic senses and these additional meanings are intended to be picked up by victims via inference. The fact that implicatures are not explicated, that is, “cannot be traced to the form of the utterance”, means that officers can avoid committing to these implicated meanings and can even deny them (denial is itself an act of gaslighting).
Findings/discussion
Following our Thematic Analysis, one key theme predominated in our corpus. That theme was “
1) questioning the victim’s experience;
2) blaming the victim;
3) favouring the perpetrator; and
4) raising the topic of the victim’s mental health.
We follow Abramson (2014) in differentiating between dismissing someone and dismissing by gaslighting. Abramson notes that “Gaslighting is [. . .] quite unlike merely dismissing someone, for dismissal simply fails to take another seriously as an interlocutor, whereas gaslighting is aimed at getting another not to take herself seriously as an interlocutor” (2014: 2). Within the theme of “
Each of the sub-categories 1–4 will be explored in more detail below.
Questioning the victim’s experience
Questioning a person’s experience is perhaps the most well-known, and therefore, a prototypical gaslighting strategy because it most conventionally captures the gaslighter’s intention to incite the gaslightee to question their reality in line with the play and film from which the concept derives. Questioning as gaslighting adheres to Stark’s (2019: 224) distinction between “reasonable argument” and “gaslighting”, as demonstrated below. It is important to note, though, that the success of gaslighting – that is, getting the gaslightee to question their reality – is not a precondition for gaslighting to have seen to have occurred. The harm results from the gaslighter “They [police] thought it [domestic abuse incident] was a joke” (R184) “police officer look[ed] me directly in the face and told me he thought I was lying” (R177) “[police asked] are you sure you’re being truthful[?]” (R118) “I was made to feel that I was making things sound worse than what they were” (R51)
In each case, the abuse was disposed of using the outcome category “No Further Action”. Specifically, R184 shared that this was despite the perpetrator in her case committing a sexual assault against her. To treat domestic abuse as “a joke” is to minimise what happened. When the officer treats serious abuse as a “joke” they are gaslighting the victim into perceiving her experience as somehow less significant than it actually was. Given the seriousness of the sexual assault, we consider the officer in R184’s case to have flouted Grice’s Quality maxim, in that he knew the assault was not a “joke” and, treating it as such, he intentionally invited the victim to question her perception of the assault. In R177’s case, the gaslighting derives from the explicit questioning of R177’s experience (“he told me he thought I was lying”). The officer goes on-record and explicitly tells R177 he disbelieves her, but he relies on implicature when communicating his rationale for his disbelief (see below, “favoured the perpetrator”). In this case the victim documented serious physical injuries that included broken bones inflicted upon her by her (now) ex-partner, yet she reports that the police “did nothing”. To compound the situation, police also
In the case of R118, she responded to the question “In your dealings with police, did you have any concerns regarding their use of language towards you or any details of your case?” by saying that police not only questioned her honesty (“are you sure you’re being truthful”), thus inviting her to question herself, but they also informed her of “the consequences [to her] of reporting” the abuse to police. Effectively, the police officer weaponised the criminal justice process by “threaten[ing]” her “with the CPS”
1
and asked her “are you sure you want to report it?” In this instance the officer has used a fear appeal (“threatened”) to dissuade the victim from progressing the case. This was followed up with an interrogative which operates as a pragmatic implicature (Grice, 1975). That is, the utterance “are you sure you want to report it” flouts the maxim of relation because the victim is indeed reporting it. Therefore, the question carries additional meaning or what Clark (2021) calls “indirectly communicated assumptions” which can be explicated as “you really don’t want to report this”. When used in the same exchange as the thematic strategy (
R51 was gaslit about the extent of the harm caused to her. She notes that she was made to feel like she was exaggerating the perpetrator’s actions (“I was made to feel that I was making things sound worse than what they were”). For context, the perpetrator choked her in a case of non-fatal strangulation. R51 wrote that the “dangerous” perpetrator had a history of violently abusing several other women. Despite this, R51 states that police “let him go”. R51’s case was also disposed of with “No Further Action”. In relation to Grice’s (1975) maxims, this demonstrates both explicit minimisation and dismissal, as well as implicature, flouting the maxim of Relation and challenging survivors as having failed to adhere to the maxim of Quality (being truthful). In short, in the examples above, the reported police behaviour exemplifies gaslighting through inviting the victims to question their reality or impose an alternative reality upon them – they either thought the victim in the case was being dishonest (“lying”) (R177 and R118), disingenuous (“a joke”) (R184), or exaggerating (R51). The implicatures are doing two primary things: a) they
Blaming the victim
This category of institutional gaslighting was a specific type of victim-blaming, wherein police are reported to have treated the victim as a) the perpetrator, or b) as bad as the perpetrator. This reinforces common gaslighting messages perpetrators communicate to victims and survivors inciting them to believe that the abuse is caused by the victim (not the perpetrator). Examples include the following: “[Police said] “We aren’t here to referee domestics” (R154) “[Police told me it was] my own fault for being on stalked on social media as I post on it” (R33) “. . . officer telling me he [the officer] had been abused by a woman. . .[and said] “women can hide things” (R180) “felt like they [police] were trying to catch me out” (R197)
In example R154, the reported utterance reframes a potentially dangerous domestic abuse “incident” as a competitive sporting activity through the metaphoric “referee”. The term “referee” undermines “domestics” (as does referring to abuse as “domestics”) by presenting them positively glossed as some kind of mutual competition in which the victim and perpetrator are equally and fairly engaged. In reality, R154 reported that such was the perpetrator’s abuse that the victim was forced to make “an emergency call to the police”. In R33, the victim is being blamed explicitly for an online stalking offence – “it’s my fault” simply because she is online. The preposterousness of the officer’s logic is the same logic that assumes if we are killed in our own homes it is our fault for being in our own homes. Both R154 and R33 are gaslit by police who reframe the abuse as being caused in part or wholly by the victim when neither was the case. Example R180 is more subtle and relies on understanding the attendant implicature to grasp its full meaning. The “woman” to whom the officer is referring when he says that “he had been abused by a woman” bears no relevance to the incident being investigated, but the victim can (and does) reasonably expect his utterance to carry relevance. This has led her to the inference that the officer is suggesting that she, too, may be abusive, just like the “woman” in the officer’s unrelated account. Indeed, the officer helps the victim’s interpretation along by making a general (and not immediately relevant) statement that “women can hide things”. The victim believes that the officer is indirectly accusing her of “hiding things” and of being “abusive” to her abuser (her abuser counter-alleged abuse, which is a common strategy of abusers). In other words, by making a generalisation about women, the officer is gaslighting the victim, inviting an implicature that paints an alternative reality than that presented by the victim (that the victim is that kind of “woman”). In R197, the victim reports that police are actively causing her to question her account by intentionally picking holes in it, “trying to catch me out”.
In these examples, and indeed in the remaining examples in this category, there is a significant shift in the attribution of responsibility regarding abusive behaviour. This chimes with research findings by Coates and Wade (2004), specifically relating to the discursive strategies employed by judges in reformulating violent (sexual) behaviour as mutual – thus obscuring the violence enacted by the perpetrator and the resistance of the victim.
Favouring the perpetrator
The third category that emerged in the data was in victims’ accounts of police praising the perpetrator or viewing perpetrators favourably. In this theme, victim accounts highlighted two key points: a) police as being partisan towards the perpetrator, and b) sharing personal opinions of the perpetrator with the victim. Favourable comments made by police that present positive characterisations of the perpetrator cannot be viewed as anything other than harmful to the victim to whom the perpetrator has been abusive. These utterances are not context-free, particularly because the hearer is the victim. As for the previous examples, these utterances gaslight by generating implicatures that cast doubt on the victims’ experience of the perpetrator. For example, R46 shared her experience of reporting abuse multiple times to the police. In some of these instances police commented on how “nice” her abusive husband “seemed”, which was contrary to her perception and lived experience. R46 also told us that the police shared a joke and laughed along with her abuser in front of her. Further examples are outlined below: “he seemed nice at interviewing” [. . .] “he’s very sorry” (R46) “You should respect him [.] [H]e’s an intelligent man” (R16) “[H]e was too nice” and “a church going, quietly spoken, respectable man” (R177)
The implicatures generated by these behaviours amount to questioning the validity of the victims’ experiences. R46’s report of the officer’s comment that her abuser “seemed nice at interviewing” was volunteered inappropriately – it is personal opinion and uttered to counter the victim’s experience of him as an abuser but without the officer going on-record in saying so. Favouring the perpetrator sends a clear message to victims and survivors; that they are the “problem” that they are not being believed. Victims and survivors reported that in making such unwarranted comments, police are potentially increasing risk rather than reducing it because victims are dissuaded from engaging with the law enforcement process. This also aligns with responses indicating a belief by victim-survivors that engagement with law enforcement made things worse, as evidenced by victims and survivors who noted that “officers are just not trained and as such empower perpetrators” (R46), “their [police] involvement actually made things worse” (36 respondents stated this), and that “I didn’t feel protected at all” (R39). Officers’ utterances to victims that indicate a favourable attitude towards the perpetrator are not remarks made in passing. They are neither appropriate nor relevant to the investigation. They cannot be interpreted in isolation. Rather, they are indicators of bias, subjective points of view, and are understood by victims and survivors to reinforce a primary message of domestic abusers – that no one will believe you.
Raising the topic of the victim’s mental health
The next theme concerns the raising of victims’ mental health by institutional personnel. The number of instances of mental health raising was low in the data – there were four instances of references to victims’ mental health by police. All four fell within the theme “Police gaslighting”. This raised a number of interesting pragmatic moves including the generation of implicatures about the state of victims’ mental capacity by asking them to report on their mental health when reporting a crime. We have termed this a mental health “paradox”, whereby victims and survivors are both stigmatised by actual and/or perceived mental health (by losing credibility as a witness as a consequence), or by its absence (by being unable to verify that they do not have poor mental health, and so losing credibility). Mental health in the discussion below becomes intertwined with perceptions of victims’ and survivors’ emotional state and, in turn, their state of mind as a consequence of the abuse. “[police] [a]sked for my mental health records” (R80) “[police] believed the perpetrator. . . [perp] told them [police] I was ‘crazy’ and had taken an overdose” (R167) “police were trying to suggest I was suffering from mental health issues instead of listening to me” (R149) “They treat[ed] me like a hysterical woman throughout and lied in their reports and when I challenged, [police] said ‘I must be confused’” (R154)
It is important to note that abusers often weaponise victims’ mental health as a defence or counter-allegation and will have gaslit victims along these lines before police get involved in a case. In the example of R80, the police are making a direct request to see the mental health records of the respondent, which presupposes that she has a mental health condition. R80 notes that she does not, in fact, have a mental health condition and that her abuser had conveyed this lie to police who appeared to accept it unquestioningly. A similar situation occurs with R167 who shares that the perpetrator’s claims of her supposed ill mental health served merely to discredit her report of domestic abuse – like R80, the allegation of ill mental health was accepted unproblematically by police. For R149, police, unprompted, called into question the state of the respondent’s mental health. Police downplayed her report of domestic abuse choosing instead to focus on her “suffering from mental health issues”. Similarly, R154 had her challenge of police misconduct dismissed through spurious claims that she was “confused” (“when I challenged, [police] said I must be confused”), which would likely have adversely impacted the validity of her domestic abuse report. R149 was able to successfully rebut the pop-psychology (Mosoff, 1995) “diagnosis” due to her being a mental health professional, but the need for her and other victims to “prove” or successfully demonstrate that they were mentally capable of making a valid claim of DA was notable in our survey responses. It is important to note that in conducting impartial investigations, police will act on information from both suspect and victim in the case and so following up mental health concerns is not an irrelevant act in terms of intelligence gathering, particularly because the CPS will often expect these claims to be followed up for rebuttal purposes in court. However, the use of ill mental health as a potential reason to disbelieve, undermine, or render implausible reports of domestic abuse raises serious concerns. Invoking the stigma around the topic of mental health more broadly, it cannot be ignored that the institutional application of female “hysteria” is present in victims’ accounts of their experience with police through cognate terms such as “crazy” (R167) (Zaccour, 2018: 67).
Another element of the mental health paradox is that perpetrators
Zaccour (2018: 70) notes in their research on the invocation of mental illness in child custody cases in Canada that “allegations [of poor mental health] shift the focus away from family violence concerns”. The same parallels can be drawn within our data, wherein the allegations of mental health concerns not only impact the lived experience of victims and survivors in the act of reporting but also represent a move away from the allegations of violence and refocus instead on the pathologising of consequential female distress. Additionally, where allegations of violence were made against the father, Zaccour (2018) found a strong correlation of mental health labels being used against the mother as a counterargument in family court.
Given the purpose of gaslighting is to alter a person’s reality and lead them to question themselves, their experiences, and their judgement, it is particularly insidious when executed by law enforcement who are tasked with protecting the public from harm. When victims are institutionally gaslit by questioning their mental state, there are serious implications. For example, being assigned a mental health label invariably leads to the involvement of mental health professionals and wider social services should the case progress, and if victims have children, this institutional intervention is often weaponised against them as leverage to withdraw charges against perpetrators. Some examples in our survey include: “[t]he threat of social services and mental health practitioners hindered [me] leaving [the abuser]” (R235). R8 had Family Court used as a weapon against her. R20 stayed with her abuser to avoid the risk of being presented in a way that would result in a “risk of losing access to my children”. Where mental health was invoked in our data, victims and survivors contended that they must first prove their mental capacity before their account would be considered. We argue that this also generates a paradox for how victims and survivors are, in effect, expected by others to present themselves. If they are upset, they are perceived as hysterical or mentally ill, a fact of which they are acutely aware (e.g. “I cried a lot and felt really stupid. My husband was calm and plausible. I felt like everyone thought I was a mess”(R112); see example R154 above); if they are calm, they are not perceived to be reacting like a “real” victim (e.g. “[victims] are unlikely to be cowering in the corner… when [police] attend [an]incident it’s unlikely that we are going to behave in a manner of which [they are going] to perceive a victim to behave – we may be angry/vocal/erratic” (R46)). Again, this appears to be a no-win situation.
In summary, our examples show that a consistent implication prevails: should victims and survivors actually have their mental health impacted as a consequence of domestic abuse, it can be used by institutional agents (such as the police, the CPS, and judges) to negatively stigmatise them. On the other hand, victims’ inability to refute spurious allegations about their mental health or, the apparent lack of externally visible signs of trauma or mental health impairments to an observer, appears to stand as proof that they are not victimised (or not victimised enough). That these pop-psychology labels are also assigned or accepted so readily by law enforcement aligns with gaslighting behaviours and reinforces the perpetrators’ abuse (as discussed below).
Reinforcing the perpetrator’s abuse
The final category in this study is “reinforcing the perpetrator’s abuse”. Unlike the previous categories, however, this is not a distinct behaviour but a consequence of other behaviours. Through the inductive coding strategy outlined above, this category was often coded in conjunction with one other of the previous categories. The impact of the previously outlined gaslighting behaviours is that in gaslighting victims and survivors, police and the justice system more widely are reinforcing the abuse enacted by domestic abusers. This finding is not based solely on our analysis of the data, but it has been explicated by respondents. Some examples of “reinforcing” behaviours are noted below: “[police said] A victim and an abuser will choose each other across a crowded room” (R147) “[police said] Call us when he actually hurts you” (R154) “[Police said] if I pursued this case that it would also affect my ex-partners job and his mental health” (R149) “[police showed a] lack of compassion. . . reinstalled what my abuser had made me feel” (R62) “[H]e [police] made me believe I was in the wrong” (R215) “The whole [time] I dealt with police made the experience much worse. I left feeling as though I had been attacked and made to feel a liar, inconvenient and that I was in the wrong” (R149)
The concept of reinforcement and gaslighting was evident in victims’ and survivors’ responses, and again, the implicatures generated by each of these statements is apparent. The extract from R147, “[a] victim and an abuser will choose each other across a crowded room”, is being made directly to a person reporting domestic abuse and raises the question of why this statement has been made. The implication is that the victim has “chosen” to be abused through some kind of active agency on their part. Similarly, “call us when he actually hurts you”, “he [police] made me believe I was in the wrong”, and, “I was made to feel a liar”, implicate that there is a threshold of abuse to be met before intervention is justified and that this threshold is exclusively physical (R154), implying that much of the psychological experiences of victims and survivors are not reflective of this external “reality” and that the abuse is not taking place (R215, R149).
Example R149, “[police said] if I pursued this case that it would also affect my ex-partners job and his mental health” reinforces the abuser’s credibility as the police stance appears to align with the perpetrator’s in this case. It is also worth noting in this extract that the mental health of the abuser is weaponised against the victim which contrasts directly with the stance taken towards the victim discussed in the examples under “mental health” above, and the mental health paradox outlined therein. It is also clear that victims and survivors are left feeling unsupported and dangerously isolated – “I felt I wouldn’t be believed or protected” (R141, speaking about “adverse police experiences”).
Institutional betrayal
When a victim calls police, rarely is it the first time the abuse has occurred, even if it is the first time the police have been called. Theoretically, engaging police who are tasked with public protection is a natural safeguarding step to take; however, this can be difficult in practice for a victim. Making that call to police often comes with great risk to the victim (Epstein et al., 2003), for example, via reprisals from the perpetrator, shame from family, threat of family separation from social services, potential homelessness, and more. Consequently, when victims make that call, it is usually a clear indicator that a victim has reached, in the words of one survivor, a “crisis” (R46). Therefore, when the institution designed to protect victims itself calls into question the validity of a victim’s experience, this constitutes a betrayal of trust that can have devastating consequences. Smith and Freyd (2013, 2014) refer to this particular breach of trust as “institutional betrayal”.
Institutional betrayal builds on studies of betrayal trauma theory by Freyd (1996) who maintains that interpersonal betrayal (between two people in some form of relationship) is more psychologically damaging than a betrayal between strangers because the trust investment in the close interpersonal relationship is (or is expected to be) greater. “Institutional betrayal”, then, occurs “when an institution causes harm to an individual who trusts or depends upon that institution” (Smith and Freyd, 2013, 2014: 578; see also Epstein and Goodman, 2019). The trauma experienced by a victim is then “exacerbate[d]” by “the action and inaction” of police (Smith and Freyd, 2014: 577). Where law enforcement (whose role is seen as “to protect”) gives a clear indication that they doubt the victim’s experience, question the victim’s mental state, align themselves with the perpetrator, or blame the victim for what has happened
Conclusion/future directions
In drawing the categories of institutional gaslighting together, a picture emerges that links the behaviours above with the concept of institutional betrayal. Police officers are representatives of their institution, whose role is to protect. Where the experiences of victims and survivors are undermined, ignored, or flipped to make victims perpetrators, the “second assault” is enacted. In reinforcing the perpetrator’s abuse, victims and survivors are caused additional trauma by the one institution that should offer them support and protection. As a result, victims are isolated further, much more reluctant to engage with institutional support networks and law enforcement and ultimately, unable to get justice.
Footnotes
Acknowledgements
The authors would like to thank all of the participants involved in sharing their experiences with us.
Ethical considerations
This project was approved by Northumbria University’s Ethics Committee (Project Ref: 2666)
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data availability statement
Due to the ethical restrictions and the sensitive nature of the data collected, this is not made available for sharing.