The Language We Use About People Who Use Violence

 

I have been in rooms where the same man was referred to as a perpetrator, an offender, a client, a participant, and a person who uses violence, sometimes within the span of a single team meeting. Nobody flagged this. Nobody appeared to notice. The words moved around him like weather, each one carrying a different implicit theory about who he was, what he had done, and what, if anything, could now be done about it.

 

That kind of unconscious drift is worth pausing on. Because the words we use in this work are not neutral containers for meaning. They do something. And what they do matters for how we engage with people, how those people engage with themselves, and, ultimately, how likely they are to change.

 

How the Field Arrived Here

 

The dominant terminology in domestic, family and sexualised violence practice has shifted considerably over the past two decades. For most of that time, the available vocabulary was stark: batterer, abuser, offender, perpetrator. These words came largely from the criminal justice system, and they carried its logic: this is a person who has committed a wrong, and the task is to name that wrong clearly and not lose sight of that.

 

The shift toward “person who uses violence” was a deliberate move by practitioners and researchers who noticed something about those earlier labels. When a word becomes a noun, it becomes an identity. A batterer is not a person who has battered someone. A batterer is a type of person. The label attaches to the self rather than to the behaviour. And once that happens, something gets foreclosed. If this is what you are, the question of who you might become is much harder to hold open.

 

Person-first language, placing the person before the description of their behaviour, has been studied in mental health, disability, and criminal justice contexts. A 2024 study (Ballou & DeWitt, 2024)found that replacing crime-first terms like “felon” with person-first language like “person with a felony conviction” produced meaningfully fewer negative associations among respondents and more support for rehabilitation. A British Journal of Psychiatry editorial published in 2026 (Tomlinson et. al. (2026) found similar effects in forensic mental health settings, noting that person-first language was associated with reduced blame, reduced stigma, and more recovery-oriented attitudes among practitioners, not just members of the public.

 

These are not just trivial effects. They describe what happens in the mind of the person using the language, not just what happens to the person being described.

 

What We Carry Into the Room

 

This matters because as practitioners we are not immune to the words in our own heads.

 

A practitioner who thinks of the man across from them as a perpetrator is, at some level, working with that frame. The word carries a set of assumptions: this person has done something, that thing defines them to a significant degree, and the work is to address that thing. That is not an unreasonable starting point for accountability work. But it can subtly foreclose the curiosity that effective practice requires. If someone is a perpetrator, their history, their ambivalence, their own experience of harm, their identity as a father or a brother or a worker, becomes secondary to the organising label.

 

“Person who uses violence” asks the practitioner to hold something more open. The violence is named clearly, as a behaviour, as a choice, as something that has caused real harm. But the person is prior to it. They are someone who has done this. They are not reducible to it. That distinction, which can sound theoretical in a training room, has practical consequences in a first session. The practitioner who genuinely holds the person as prior to the behaviour will ask different questions, listen differently, and notice different things.

 

Victoria’s MARAM practice guidance makes the point plainly. When engaging with practitioners, the guidance recommends using “person using family violence” or “person who uses violence” in direct practice. “Perpetrator,” it notes, should be reserved for legal and policy contexts. The reason is not to soften accountability. It is because a practitioner who calls the man across the table a perpetrator, to his face or in their own framing of the work, is doing something to the relationship before the conversation has properly started.

 

The Accountability Problem

 

The strongest argument against this direction comes from the family violence sector, and it deserves to be taken seriously.

 

Terminology that softens the description of violent behaviour can, in some contexts, soften accountability for it. Some practitioners working with victim-survivors have raised concerns that “person who uses violence” sounds like a description of someone who happened to pick up violence somewhere, like a tool lying in the road, rather than someone who made repeated choices to harm a partner and children. There is something in this. Language shapes narrative, and narrative shapes responsibility.

 

The further concern is about weaponisation. Some men who use violence are sophisticated observers of the language used by the services working with them. Sector jargon can be learned and turned back. A man who has absorbed the vocabulary of trauma, coercive control, and misidentification of the primary aggressor can use those terms to cloud accountability conversations rather than support them. The words that were designed to centre the experiences of people experiencing violence can be co-opted into a defence.

 

These are real risks. They argue not against person-first language, but for precision and care in how it is used, and in what settings. There is a meaningful difference between the language a practitioner uses in their own framing of the work, which should centre the person’s capacity to change, and the language used in a legal report, a risk assessment, or documentation shared with partner agencies, which needs to be unambiguous about what has happened and who is responsible for it.

 

The Desistance Argument

 

The most compelling case for careful language, in my view, comes not from the terminology debate itself but from the research on desistance, on how people who have caused harm actually stop doing so.

 

A consistent finding across desistance research is that identity change is central to sustained behaviour change. People who stop using violence, who stop offending, who stop using substances, tend to describe a shift in how they understand themselves. They develop what researchers call a desisting narrative identity: a story about who they are that makes the previous behaviour inconsistent with who they want to be. The work of a good program, in this framing, is not just skill-building or accountability in the narrow sense. It is supporting a person to construct a different account of themselves.

 

Labels that fix identity make that process harder. If a man has been told, repeatedly and by multiple systems, that he is an offender, a perpetrator, a batterer, the work of building a different self-narrative starts from a significant deficit. The label he has been given does not leave much room for who he might be becoming. Person-first language does not resolve this on its own. But it holds open the possibility that the work is about becoming, not just about being held accountable for having been.

 

What This Looks Like in Practice

 

None of this is an argument for avoiding the behaviour in conversation. The most useful language with someone who has used violence is often quite specific and direct: what happened, what the impact was on the people in the room, what choices were made. Euphemism serves no one. Neither does confrontational labelling.

 

What it looks like, in the room, is something closer to: separating the behaviour from the person consistently, naming the behaviour clearly when naming it matters, using curiosity about the person’s own account of who they are and who they want to be, and not allowing the label on the referral to set the ceiling on what is possible.

 

That is not a soft approach. It is a demanding one. It requires us as practitioners to hold both things simultaneously: the reality and impact of harm that has been done, and the genuine openness that change is possible. Language is one of the tools that helps them do that, or makes it harder.

 

A Final Word

 

The words we use in this work travel. They travel from intake forms to team meetings, from case notes to court reports, from one practitioner to the next. They also travel inward, shaping how as practitioners we think about the people sitting across from us, and, in some cases, how those people think about themselves.

 

That is worth taking seriously, not as a reason to avoid naming harm clearly, but as a reason to be deliberate about the words we choose and what work we are asking them to do.

 

The man who was called five different things in that meeting deserved at least one of them to be about who he is, not just what he has done.

 

 

Denver, M., Ballou, A., & DeWitt, S. E. (2024). What’s in a label? Public use and perceptions of labeling alternatives in criminology. Justice Quarterly, 41(6), 763–789. https://doi.org/10.1080/07418825.2024.2332373

Tomlin, J., Mitisheva, A., Cornetchi, N., & Kilbane, S. (2026). The value of person-first language in forensic mental health services. BJPsych Bulletin, First View, pp. 1–4. https://doi.org/10.1192/bjb.2026.10230

 

 

Published on Wednesday, July 15th, 2026, under Family violence, Programme design & development, What Ken thinks

---

Comments are closed.