Gaslighting

One-liner: Manipulating someone into doubting their own perceptions, memory, or sanity.

Also known as / related terms: Reality distortion, “crazymaking,” coercive control (adjacent), mind games.

What it is

It makes you doubt what you saw.

Gaslighting is a pattern in which one person repeatedly denies, contradicts, or reframes another person’s directly experienced reality until the target starts to distrust their own memory and judgment. The APA Dictionary of Psychology defines it as manipulating another person into doubting their perceptions, experiences, or understanding of events, and it notes the term historically described manipulation so extreme it could justify psychiatric commitment, though it’s now used more broadly. The name comes from the 1938 stage play Gas Light (and its 1940/1944 film adaptations), in which a husband dims the gaslights and then insists his wife is imagining it. Clinically, “gaslighting” is not a diagnosis itself. Psychologists note it is a colloquial description of a manipulation pattern, sometimes associated with antisocial or narcissistic traits, rather than a disorder in its own right. Academic work on workplace gaslighting treats it as a distinct organizational construct: a pattern of behavior, embedded in a power-unequal relationship, that over time erodes a target’s confidence in their own perception of events.

Gaslighting exploits a basic, well-documented fact about how memory actually functions. Elizabeth Loftus and John Palmer’s 1974 experiment found that simply changing the verb in a question about a car crash, asking how fast the cars were going when they “smashed” versus when they “hit”, changed witnesses’ recalled speed estimates, and led some witnesses to falsely remember broken glass that was never there. Loftus’s later synthesis of three decades of this research, “Planting Misinformation in the Human Mind” (2005), confirmed the general finding: memory is not a fixed recording that gets played back, it is reconstructed each time it’s retrieved, and it is genuinely vulnerable to being altered by information encountered after the fact. A confident, repeated denial is exactly this kind of post-event information. This is also why the standard defense, a contemporaneous written record, works as well as it does: Hardin and Higgins’s shared reality theory (covered in more depth in this guide’s Covert Digs / “Dog Whistling” entry) holds that people confirm their own perceptions through social verification with others. A dated email is a form of external memory the misinformation effect cannot silently rewrite the way it can a purely internal recollection, which is precisely why gaslighters push so hard against paper trails specifically.

What it looks like (workplace)

Before: An employee with an ordinary, healthy level of confidence in their own read of things. Small workplace disputes come and go and get shrugged off without much residue.

The incident: In a one-on-one, a manager flatly denies a project deadline was ever discussed, despite a dated email thread proving otherwise. When the employee points to the email, the manager says, “You’re remembering it wrong, you always do this.”

The month after: The employee starts pre-writing recap emails after every conversation with this manager, just in case. They find themselves re-reading their own sent mail before trusting what they remember saying. More than once, they’ve turned to a colleague after a meeting and asked, quietly, “did I read that right?” This is deliberately the same image as the recap-email habit described in this guide’s Dark Triad entry, and the resonance is intentional: it is the same reconstructive-memory vulnerability, playing out the same way, whether the pattern comes from a Dark Triad-consistent operator or from someone simply willing to deny an inconvenient fact.

How it actually unfolds

1. Small, deniable contradictions. Testing deference at low stakes, “that’s not what I said,” about something minor enough that pushing back feels disproportionate.
2. Escalation to pattern-claims. “You always do this” moves the target from questioning one memory to questioning their memory in general, combining Labeling with a direct attack on recall.
3. Isolation from verification. Reality-checks with colleagues get discouraged or quietly made harder, cutting the target off from the social verification shared reality theory says memory depends on; this often runs alongside Workplace Ostracism / Social Exclusion.
4. The payoff: outsourced reality. The target now checks their own perception against the manipulator’s account by default, and the burden of proof has fully flipped, from “prove this happened” to “prove you’re not misremembering.” If the target ever produces hard evidence, expect DARVO’s Reverse stage next.

Stage 1 costs the target a small, dismissable moment of doubt. Stage 2 costs them confidence in their memory generally, not just about this one event. Stage 3 costs them the outside checks that would normally catch the distortion. Stage 4 costs them their own authority over their own experience.

Why they do it

It shifts the burden of proof onto the target. By making their version of events the only “sane” one, the person avoids accountability and consolidates control over the narrative, and because memory really is reconstructive rather than archival, a confident enough denial has genuine psychological leverage to work with, not just social pressure.

How to protect yourself

Your memory is not the unreliable one here.

• Keep a contemporaneous written record (email recaps, dated notes), a paper trail counters the “you’re misremembering” move because it’s external memory the misinformation effect can’t reach; see Document Contemporaneously.
• Say the fact plainly and stop: “The email is dated March 3rd. I’m not going to debate what it says.” Don’t re-argue the underlying reality.
• Get a second read from a trusted, uninvolved witness before assuming your perception is off, this restores the social verification the isolation stage tries to cut off.
• Notice the internal signal: chronic self-doubt after interactions with one specific person is itself data.
• If it’s a pattern, name it privately in writing (to yourself or HR) using neutral, factual language, not accusatory labels; see Getting Your Own Judgment Back for rebuilding trust in your own read after a sustained pattern.

Cross-links: DARVO, the reversal that often follows once hard evidence appears; Reframing, the frame-swap engine that runs alongside outright denial; Covert Digs / “Dog Whistling”, self-doubt manufactured through a single loaded phrase instead of denial; Labeling, the trait vocabulary behind “you always do this”; Moving the Goalposts, reality-shifting applied to success criteria rather than memory; Plausible Deniability, the cover that protects each individual denial; Document Contemporaneously, the external-memory defense; Getting Your Own Judgment Back, the recovery path.

Sources:

Label note: Colloquial term describing a real manipulation pattern; not itself a clinical diagnosis, though referenced in clinical and organizational-research literature. The misinformation-effect and shared-reality research cited here are genuine, separately-established academic constructs, applied to gaslighting as an explanatory mechanism rather than studies of gaslighting itself.