Case Study 23-02: DARVO — Research, Recognition, and Application

Overview

DARVO — Deny, Attack, Reverse Victim and Offender — is not a colloquial term invented by conflict coaches or self-help writers. It is a clinically observed and empirically studied pattern, first named and described by psychologist Jennifer Freyd in 1997, drawn from her research on betrayal trauma and institutional responses to abuse. This case study examines the research behind DARVO, the psychological mechanisms that drive it, and what empirical work tells us about how to recognize it in real-time conflict — not just in high-stakes abuse disclosures, but in the ordinary difficult conversations of everyday life.

Jennifer Freyd and the Origins of DARVO

Jennifer Freyd is a research psychologist whose primary work centers on betrayal trauma theory — the psychological effects of trauma inflicted by trusted others, particularly when institutions protect perpetrators over victims. Her observation of DARVO emerged from examining how people and institutions respond when accused of wrongdoing, particularly sexual abuse.

In her 1997 paper and subsequent work, Freyd noted a strikingly consistent pattern in how accused parties responded to allegations:

First: Deny. The accused party flatly denies the behavior occurred. This denial is typically not qualified or conditional — not "I don't remember it that way" or "I think you may have misunderstood" — but absolute. "That never happened." "That is not what occurred."

Second: Attack. The accused party attacks the credibility, character, mental stability, or motives of the person raising the concern. This attack is often wide-ranging: the accuser is described as unstable, motivated by grievance, confused, or dishonest.

Third: Reverse Victim and Offender. The accused party re-positions themselves as the true victim — of the false accusation, of the unfair treatment, of the damage being done to their reputation and wellbeing. The person who raised the concern is now the aggressor.

What made Freyd's observation significant was not just that these behaviors occurred, but that they occurred in a consistent sequence, operated at both individual and institutional levels, and had measurable effects on the people who experienced them.

The Psychological Mechanisms Behind DARVO

Understanding why DARVO works requires understanding several intersecting psychological processes.

Why Denial Is Effective: The Authority Effect of Certainty

Human beings are wired to weight confident assertions more heavily than uncertain ones. Research on eyewitness testimony (Loftus, 1979; Wells & Bradfield, 1998) has consistently shown that the confidence with which a memory is held has very little relationship to its accuracy — but that confidence strongly influences how much others believe the person reporting it.

When an accused party delivers an absolute, confident denial — "That never happened" — the person who raised the concern faces a genuine epistemological problem: they have their memory of the event, and the other party has their unqualified denial. For the accuser, particularly in cases where they lack independent corroboration, the confident denial can create genuine self-doubt. The question shifts from "did this happen?" to "did I remember it correctly?" And that shift is the beginning of the denial's work.

Freyd's research noted that this effect is amplified in relationships of trust or dependency. When the person denying is someone whose competence, judgment, or authority we have reason to respect — a supervisor, a parent, a partner — their confident denial carries even more weight. Our relational history with them is effectively offered as evidence against our own perception.

Why Attack Works: The Credibility Game

Once denial has introduced uncertainty about the accuser's memory or perception, the attack step capitalizes on that uncertainty. The attack typically takes the form of character evidence: the accuser is presented as someone whose claims should not be trusted, for reasons that reach beyond this particular incident.

Research on how juries evaluate witnesses and how organizations handle internal complaints (Crenshaw, 1991; Poepsel & Rosenbaum, 2011) shows that credibility attacks are effective precisely because they are often unfalsifiable. "She has a history of overreacting" cannot be proved or disproved in a single conversation. "He's always looking for a grievance" is an assertion about a pattern rather than a fact. These claims function not by proving the accuser is wrong but by making it feel risky to take their side — because if the attack is even partially right, everyone who believed the accuser is implicated.

The attack also serves a second function: it produces a new emotional event in the accuser. Being attacked — having your character, stability, or honesty questioned — triggers its own threat response. The accuser is now dealing not only with the original concern they raised but with the fresh wound of being characterized as unstable or dishonest. This emotional load is additional to whatever emotional weight they were already carrying, and it reduces their capacity to stay coherent and on-topic.

Why the Reversal Is the Most Powerful Step: The Exploitation of Empathy

The Reverse Victim and Offender step works by exploiting one of the most prosocial instincts human beings have: the impulse to relieve suffering. When someone appears to be in pain — when their distress is visible and visceral — we feel the pull to make it stop. And in the reversal, the accused party makes their distress visible and visceral.

The implicit logic of the reversal is that the accuser has caused the accused party's suffering by raising the concern. The suffering is real: being accused of wrongdoing is painful, regardless of whether the accusation is accurate. The manipulation lies not in manufacturing the feeling but in weaponizing it — using it to make the accuser responsible for the accused's pain, and by extension, responsible for ending it by dropping the concern.

What makes this particularly effective is that the empathic pull toward someone in pain is not a weakness or a flaw. It is one of the fundamental mechanisms of social cohesion. DARVO exploits a virtue. The person who continues to press their concern despite another person's visible distress can feel — from the inside — like a villain, even when their concern is entirely legitimate.

Freyd's research found that this step was particularly effective in contexts where the accuser already had some guilt or ambivalence about raising the concern — which, in most difficult conversations, is exactly the context. Nobody arrives at a confrontation feeling entirely confident. DARVO finds the existing doubt and expands it.

DARVO at the Institutional Level

One of Freyd's most significant contributions was documenting that DARVO operates not only at the individual level but at the institutional level. Organizations, departments, governments, and systems can collectively enact DARVO in response to disclosures of wrongdoing.

Institutional denial: "Our review found no evidence of wrongdoing." "The process was followed correctly." "There is no pattern here."

Institutional attack: "This is a personnel matter." "The person raising this concern has a history of complaints." "We have concerns about the accuracy of this account."

Institutional reversal: "These unfounded allegations are damaging the institution and its employees." "We are the ones being harmed by this campaign." "The real victims here are the staff who are being unfairly accused."

Research by Smith and Freyd (2013, 2014) documented that institutional DARVO is not only common but often more effective than individual DARVO — because the institution has significantly more resources for the Deny and Attack steps (legal teams, communications departments, reputational weight), and because the institutional reversal can claim that a broad community of people — staff, constituents, members — are being harmed by the accuser's claims.

For students of conflict, this has a practical implication: DARVO is not a tactic used only by unusually manipulative individuals. It is a pattern available to — and frequently used by — people and systems that are otherwise respected and functional, when they feel threatened by a legitimate concern.

Recognizing DARVO in Real-Time

The research literature suggests several signature markers of DARVO that can be detected in real-time, even within a conversation that is moving quickly:

The direction of the conversation has reversed. You began with a concern about something someone did. The conversation has moved to something you did — specifically, something about the act of raising the concern. This reversal is the most reliable indicator.

The flat denial. There is a qualitative difference between "I remember it differently" (a genuine memory dispute) and "That never happened" (DARVO denial). The latter has a categorical quality — not a claim about memory but a claim about fact, delivered with certainty.

The scope of the attack is disproportionate. The attack does not address the specific claim but ranges widely across your character, history, motives, or mental state. A response to "I felt dismissed in our meeting" that describes your entire pattern of behavior over two years is disproportionate in a way that signals the attack function.

The accused party's suffering is centered. The emotional weight of the conversation has shifted from your experience (which prompted the conversation) to their suffering (caused by your accusation). You are now managing their pain rather than articulating your concern.

You feel responsible for ending the conversation. If you feel like continuing the conversation makes you cruel — if stopping and apologizing feels like the moral thing to do, despite having raised a legitimate concern — this is the reversal operating.

The "Day After" Problem: Why DARVO's Effects Outlast the Conversation

One of the more troubling findings in the DARVO literature concerns what Freyd and colleagues call the "day after" effect: the pattern's most powerful impact is often not immediate but delayed. During the conversation itself, the accuser may maintain their position. But in the hours and days following, with Harmon's confident denial and condescending attack replaying in memory, with the image of his distress at the reversal still vivid, many people begin to wonder if they were wrong.

This is not a sign of weakness or instability. It is the product of the same psychological mechanisms that make the tactic effective in the first place: the authority weight of confident denial, the credibility damage of the attack, and the empathic pull of someone's expressed suffering. These mechanisms continue operating in memory long after the conversation has ended.

The practical implication: people who have experienced DARVO should be expected to feel more doubt after the conversation than during it. Having a written record of the original concern — made before the conversation — and a trusted person with whom to debrief afterward provides a crucial check on memory revisionism. Not because the person who experienced DARVO is unreliable, but because the tactic is specifically designed to make reliable perceivers doubt their reliability.

DARVO in Everyday Conflict

Freyd's original research focused on serious abuse disclosures. Subsequent application by conflict researchers and practitioners has identified DARVO operating at lower stakes — in family disputes, workplace grievances, friendship conflicts, and ordinary relationship conversations.

The mechanisms are the same. The scale differs. A parent who responds to a child's complaint about a harsh criticism with "I never said anything like that," followed by "you've always been oversensitive," followed by "this is very hurtful — I can't believe you think I'm a bad parent," is deploying DARVO at low intensity. The conversation that results is recognizable: the child is now comforting the parent, the original criticism has not been discussed, and the child is uncertain whether they had a right to raise the concern.

This does not mean every denial-attack-reversal sequence is deliberate or calculated. One of the most important findings in the applied literature is that DARVO can be enacted by people who have no conscious awareness that they are doing it. The pattern is available to human threat-response systems regardless of intent: when cornered, deny; when pressed, attack; when cornered further, make your pain the center. These are not strategies learned in a manual. They are deeply available responses to the feeling of being accused.

This matters for the student of difficult conversations, because it means that recognizing DARVO should not automatically lead to attributing sinister intent. The more productive question is not "are they doing this on purpose?" but "is this pattern preventing the legitimate concern from being addressed?" — and if so, what can be done to address it anyway.

Research Snapshot: Key Findings

The bystander finding deserves particular attention: the reversal step is so effective that even people who observed the full conversation from the outside — who saw the original concern, the denial, and the attack — are likely to side with the accused party after the reversal. This is a profoundly sobering data point for anyone who believes that having witnesses present protects against DARVO. It does not, without explicit preparation.

Practical Implications for Difficult Conversations

For anyone preparing to raise a concern in a context where DARVO might be deployed:

Document before. Write your concern down, with specific details, before the conversation. Date it. This creates a pre-conversation record that your memory cannot retroactively revise.

Bring a trusted second. In high-stakes contexts, having someone present who is there explicitly to observe — not to advocate — provides a check on the reversal's effects. Brief them beforehand on what DARVO is.

Prepare for the day-after doubt. Know, going in, that you may feel more uncertain tomorrow than you do today. This is not because you were wrong. It is because the tactic is designed to produce that effect.

Name the process, not the person. "I notice that we've moved from talking about [behavior] to talking about whether I had a right to raise it" is more productive than "you are using DARVO on me."

Return to specifics. Every iteration of every DARVO component expands into generality. The denial generalizes ("that never happens"). The attack generalizes ("you're always doing this"). The reversal generalizes ("this is so damaging to me"). The antidote is always to narrow back to the specific: what happened, on what day, with what outcome.

Conclusion

DARVO is not an exotic tactic used only by abusers or manipulators. It is a pattern that emerges, often unconsciously, wherever people feel accused of something — from the most serious institutional failures to the most ordinary relationship conflicts. Its effectiveness is documented. Its mechanisms are understood. And its power lies precisely in its exploitation of virtues — empathy, fairness, the benefit of the doubt — that are otherwise assets in difficult conversations.

The goal is not to become suspicious of everyone who expresses pain or denies a claim. The goal is to recognize the sequence, hold onto the specific concern that prompted the conversation, and find a way back to the question that started everything: not "who is the real victim here?" but "what happened, and what do we do about it?"

Reflection Questions

1. Freyd found that DARVO can operate unconsciously — without deliberate intent. What are the ethical implications of confronting someone for a pattern they may not be aware of?

2. The bystander study found that even witnesses who saw the full exchange tend to side with the accused after the reversal step. What does this tell us about the limits of "just have someone else present"?

3. How might cultural context influence which component of DARVO is most effective? (Consider cultures with high power-distance, or cultures where expressing personal suffering is more or less acceptable.)

4. If DARVO can be unconscious, is it possible to inadvertently use it yourself? Have you ever responded to an accusation with a denial-attack-reversal sequence without strategic intent? What were the circumstances?

5. The "document before" recommendation assumes that the concern can be anticipated in advance. What do you do when you have not anticipated the accusation — when someone raises a concern about you that you did not see coming, and you feel the pull toward denial-attack-reversal in real time?