Your manager tells you in a 1:1 that your project approach is solid. Two weeks later in a team meeting, they say: “I’ve had concerns about this approach from the beginning.” You start to speak up - you remember that conversation clearly - but before you finish your first sentence, they cut in: “I think you’re conflating my support for the team with endorsement of the specific technical direction.”
You have notes from the 1:1. You know what was said but now you’re in a room with eight people who just heard your manager’s version delivered with complete certainty. Your reality and theirs don’t match. The meeting moves on. Later, when you check the meeting notes your manager wrote, their version is what gets documented.
This is workplace gaslighting operating through hierarchical power. The linguistic patterns from Part 1 - the three types of assertives, the double direction of fit, the face-threatening acts disguised as concern - all function differently when deployed by someone with institutional authority. And the reality-distorting effects compound over time in ways that reshape organisational memory and career trajectories.
How The Patterns Change With Power
Part 1 outlined the core behavioural patterns: outright lying and denial, reality manipulation and questioning, scapegoating and blame deflection, trivialising and minimising. In hierarchical contexts, these acquire institutional weight.
When your manager says “I never approved that timeline,” they’re not just making a claim about the past. They’re exercising definitional authority, the power to determine what counts as true within the organisation. Their explicit assertive carries institutional credibility that yours doesn’t.
Performance reviews deploy covert assertives: “While Sarah believes they demonstrated leadership, peer feedback suggests a more collaborative approach is needed.” That embeds the claim that Sarah’s self-assessment is wrong inside professional language. Sarah’s manager might not even recognise it as manipulative, it’s how reviews are written.
As I explored in Why Some People’s Stories Always Get Believed, this is linguistic capital in action. Senior people spend from abundance. They can make reality-contradicting claims with minimal social cost. You spend capital you don’t have. Even with evidence, pushing back becomes a face-threatening act that violates norms about deference.
The Compounding Effect
Individual instances are destabilising. The real damage comes from how these patterns compound.
With the first instance: Manager contradicts something you clearly remember. You don’t push back. Maybe you misunderstood.
By the third instance: You start keeping notes - something feels wrong.
And then by the tenth instance: They’re so certain. Other people accept their framing. Maybe you are forgetting these conversations?
This is epistemic uncertainty taking hold, that fundamental doubt about your ability to know what’s real. Each instance makes you slightly less confident. Over time, their confident assertions override your increasingly uncertain memory.
As I discussed in The Meeting Narrative Battle, when the person who controls narrative framing also controls written records, their version becomes organisational reality. The gaslighter’s reality doesn’t just dominate in the moment, it becomes the historical record.
Meeting happens. You present strategy. Manager seems supportive. One week later: “concerns I raised about viability.” One month later, meeting notes say: “Leadership raised questions about viability.” Six months later: “Despite initial concerns...” or “As anticipated, the flaws identified early...”
The original support vanishes from organisational memory. The retrospective framing becomes what happened.
Why “Speaking Up” Advice Fails
The standard advice:
“Just speak up.”
“Just be more confident.”
“Just advocate for yourself.”
That word—”just”—does linguistic work. Linguistic research shows “just” functions as a minimiser and mitigator. When someone says “just speak up,” they’re using “just” to downplay the difficulty. The word signals the speaker views this as simple, straightforward, requiring minimal effort. I’ve learnt to never trust advice that starts with “just”.
This reflects speaker bias, the minimising actions that feel easy while underestimating difficulty for someone in a different structural position. The person giving advice likely hasn’t operated without definitional authority. For them, speaking up carried minimal cost. The “just” encodes their experience as universal but hierarchical gaslighting creates conditions where speaking up carries enormous cost:
You’re contradicting someone with definitional authority. They’re not offering a perspective, they’re exercising institutional power to determine organisational reality. Your contradiction challenges their authority with professional consequences.
You’re violating conversational norms about deference. Turn-taking research shows interrupting or contradicting someone higher in hierarchy is heard as more transgressive than the same behaviour between peers.
You carry the evidential burden. Senior claims get accepted with minimal proof. Your claims require extensive evidence to be considered. When you speak up, you must justify your reality in ways they never have to justify theirs.
The system interprets your pushback as the problem. Linguistic capital research shows identical behaviours carry different meanings based on who performs them. Senior people pushing back firmly are “strong leaders.” Junior people doing the same are “difficult” or “not coachable.”
Your documentation doesn’t necessarily protect you. You present emails and notes. The response: “You’re interpreting these through a defensive lens” or “Your notes reflect your perception, which has been affected by stress.” Evidence gets dismissed because you’re presenting it.
”Just speaking up” assumes roughly equal standing, assumes your version will be heard as credible, assumes contradicting carries similar cost regardless of position. None hold in hierarchical gaslighting.
Counter-Strategies That Work Within Constraints
There’s no silver bullet. Hierarchical gaslighting is hard to spot and harder to articulate to someone who hasn’t experienced it. If you’re thinking about sharing these articles with HR, go ahead - though explaining linguistic patterns to people unfamiliar with discourse analysis often lands as “overanalysing” rather than pattern recognition which is, honestly, such a shame.
What you can do is develop awareness that protects your reality even when you can’t change theirs.
Recognising Patterns So You Don’t Internalise Them
The first protection is seeing manipulation as it happens. When you recognise covert assertives, inclusive assertives, face-threatening acts disguised as concern, reality contradictions without repair, you stop treating these as evidence you’re confused, forgetful, or incompetent.
”I’m worried you’re not tracking things clearly” becomes visible as an attack on competence wrapped in care language, not genuine concern.
”When you’re less defensive, we can discuss what actually happened” reveals itself as an inclusive assertive embedding two claims: you’re currently defensive, and your version is wrong. Neither might be true.
This doesn’t solve the problem but it changes where you locate the problem. The issue is manipulative discourse structure, not your failing memory. That shift matters for maintaining epistemic confidence - trust in your own knowledge - even when someone with institutional power contradicts you.
Protecting Your Sense of Reality
When someone with definitional authority rewrites what happened, you need mechanisms to maintain confidence in your perception without requiring their validation.
Contemporary documentation: Note what happened close to the event. Not for HR (though it might serve that function later), but for you. When they say “I never approved that timeline” three weeks later, you have notes from that day saying “timeline approved in 1:1.”
This is for your reality, not theirs. As Bette Ludwig writes, AI can help - this is one of my favourite articles. Dictate what happened after significant interactions and let AI organise it. The point isn’t building a legal case. The point is having an external record that confirms your perception when someone contradicts it.
Pattern tracking: Once is an incident. Twice is coincidence. Five times is a pattern. Track how often this person contradicts previous statements, attacks your perception while seeming concerned, blames you for their failures. Pattern recognition helps you see the systematic nature rather than treating each instance as isolated confusion.
Trusted external reality checks: Find someone outside the situation - not in your reporting line, ideally not in your organisation - who can serve as a reality anchor. When you doubt whether the manager actually said X or whether you’re being too sensitive, you have someone who can confirm “no, that behaviour is manipulative” or “yes, your documentation shows contradiction.”
This person can’t fix the situation but they help you maintain confidence that your reality is accurate when organisational dynamics push you to doubt it.
Knowing When to Document vs. Disengage vs. Exit
Different situations call for different strategies, all acknowledging you can’t change someone with more definitional authority.
Documentation mode: When this might escalate to HR or legal issues, when you need a paper trail for performance reviews, when the gaslighter controls official records. Build clear, logical documentation that’s harder to dismiss. You’re creating the alternative record of what actually happened.
Edith Podhovnik, PhD gave me a great link for a documentation structure for these use cases.
Disengagement mode: When engaging gives more opportunities for manipulation. Stop trying to correct the record. Stop presenting evidence. Do your work, document for yourself, minimise interactions where they can deploy these patterns. You’re not changing their reality. You’re protecting yours by reducing exposure.
Exit planning mode: When staying costs exceed any benefit, when your reality has eroded so you’re questioning basic perceptions, when gaslighting has made the environment psychologically unsafe. This isn’t giving up. This is recognising some situations can’t be fixed from your position, and protecting yourself means leaving.
The hardest part is that these strategies involve accepting limits on what you can accomplish against institutional power. You can’t make them stop. You can’t make others see what they’re doing. You often can’t even name it without sounding like you’re the problem.
Understanding What You Can and Can’t Control
You can control your documentation and reality tracking. You can control who you talk to. You can control how much energy you invest in correcting their version versus protecting your own. You can control your exit timeline.
You cannot control their behaviour. You cannot control how others interpret your pushback. You cannot control whose reality becomes organisational truth when you lack definitional authority. You cannot control whether HR understands linguistic manipulation patterns.
The linguistic analysis from this series reveals why these limits exist. Gaslighting exploits power asymmetries that are structural, not individual. Your awareness of covert assertives and reality-replacement patterns is valuable but awareness doesn’t redistribute institutional power.
What it does do is help you see clearly what’s happening, maintain your grasp on reality despite sustained contradiction, make informed choices about where to invest energy. Sometimes that means strategic documentation. Sometimes calculated disengagement. Sometimes planned exit. (Some great advice from Sophia Rook ♟️ below).
None of these are defeat. They’re recognition that when someone with definitional authority systematically distorts reality, your job isn’t to fix them or make others see what you see. Your job is to protect your reality and make choices about whether staying serves you.
The Metacognitive Shift
The most powerful counter-strategy is metacognitive awareness - thinking about how you think and communicate in these situations.
You start noticing: “They just deployed a covert assertive. They’re embedding an attack on my credibility inside a question.” Or: “This is the third time they’ve contradicted documentation I have. This is a pattern, not confusion.”
This awareness means you’re no longer purely inside the experience of being gaslit. You’re also observing the linguistic mechanisms. That observer position - even when you can’t change the situation - provides psychological protection. You maintain dual awareness: the emotional reality of being undermined, and the analytical understanding of how the undermining works structurally.
People who develop this metacognitive capacity can feel more in control even within hierarchies they can’t change. You’re not eliminating the manipulation. You’re refusing to internalise it as truth about yourself. You’re seeing the discourse patterns as patterns, which means seeing them as things being done to you rather than revelations about your inadequacy.
This is the awareness I keep returning to: when you understand the linguistics, you can’t be made to believe their version over your own in the same way. You might have to perform acceptance for professional survival but internally, you know what’s happening. You know it’s manipulation, not your failing memory or questionable competence.
That knowledge - even when it doesn’t change the external situation - changes your internal experience of it. Which sometimes is the only thing you can control.
When Organisations Lose Signal
Beyond individual harm, organisations that allow senior people to gaslight create systematic problems with information flow.
When junior employees learn their reality doesn’t count, they stop raising concerns. When people who spot problems know their observations will be questioned or rewritten, they stay quiet. The organisation loses signal it needs from people closest to the work.
This connects to The Meeting Narrative Battle - when interruption patterns and narrative control track power rather than expertise, organisations filter out information from people who might see what those with power miss. Expensive failures often come from powerful people who couldn’t hear warnings from less powerful ones.
The gaslighting patterns examined in this series don’t just harm individuals. They damage organisational capacity to process accurate information. When the person who controls official records deploys reality-distorting language, organisational memory becomes unreliable. When definitional authority systematically contradicts evidence, the organisation’s grasp on what happened erodes.
As we’ve seen from AI, these patterns get reproduced, circulated, and normalised since it’s the discourse model for how LLMs are trained. People read AI-generated text that uses gaslighting structures and internalise those structures as “professional communication.” They start using the same patterns themselves. The discourse structures that make gaslighting effective spread through organisational communication like wild fire.
What This Series Revealed
Part 1 examined linguistic structures: speech acts with double direction of fit, three types of assertives, face-threatening acts disguised as concern, anti-performatives. We learned why there’s no checklist (welp) - gaslighting mimics cooperative communication while violating it.
Part 2 showed how AI systems reproduce these patterns through training on human discourse, creating direct gaslighting and discourse reproduction. Automated systems scale the linguistic patterns that destabilise reality.
This piece examined how hierarchical power amplifies effectiveness. Linguistic capital determines who deploys reality-contradicting claims with minimal cost. Definitional authority determines whose version becomes organisational truth. Documentation often serves gaslighters rather than protecting targets.
These patterns operate through language structure, not individual malice. Understanding linguistics reveals how manipulation works at discourse level. That understanding doesn’t solve power asymmetries, but it makes mechanisms visible.
You can see covert assertives attacking credibility. You can recognise inclusive assertives embedding inadequacy claims. You can track reality contradictions without repair. You can notice anti-performatives performing the opposite of what they claim.
Seeing patterns doesn’t redistribute linguistic capital but it protects your reality by showing you the problem is discourse structure, not your perception. When someone with definitional authority contradicts you, you can maintain epistemic confidence by recognising linguistic manipulation rather than internalising doubt.
The hardest truth: awareness has limits. You can see patterns, document reality, protect perception - and still operate where your version doesn’t count and theirs becomes organisational truth. Sometimes the most realistic strategy is recognising when a situation can’t be fixed from your position and choosing to leave rather than continuing to have your reality systematically undermined. (I’m including my #1 favourite article from this year here).
The linguistic analysis makes manipulation visible. What you do with that visibility depends on your situation, your resources, and your assessment of whether the environment can change or whether protecting yourself means leaving it.
This is Part 3 of a three-part series on gaslighting and linguistics. Part 1 examined the linguistic structures of human gaslighting. Part 2 explored how AI systems reproduce these patterns. This piece examined how hierarchical power amplifies gaslighting in workplace settings and offered strategies for maintaining reality when someone with definitional authority systematically distorts it.
