How do I know if I'm being emotionally manipulated?

You know you are being emotionally manipulated less by any single moment than by a pattern: disagreements that keep ending in doubt about your memory, silences that only lift when you give in, and a life with gradually fewer choices in it. Everyone tries to influence a partner, so the more useful test is whether that influence works by persuading you or by making you less sure of yourself and less free.

What do researchers actually mean by manipulation?

Start with the uncomfortable part. Nearly everyone manipulates a little. In a 1987 study, psychologist David Buss and colleagues catalogued the tactics people use to get someone to do something or stop doing it, and found six families: charm, the silent treatment, coercion, reason, regression and debasement.1 Charm was the go-to for getting someone to start something; coercion and the silent treatment were used more to make them stop.1

In plain terms: a sulk on a Sunday is not a diagnosis. The word “manipulation” becomes useful when a tactic stops being an occasional move and becomes the way decisions get made, and when what it targets is your confidence in your own perception.

Consider Rania and Gideon, a year together. Rania remembers him agreeing to her sister’s birthday dinner on Saturday. On Friday he says he never agreed, that she “always does this” and that she has been forgetful lately. She scrolls back and finds his text: “sure.” He says that meant something else. For the next two days his replies are one word long, until she apologizes.

What is gaslighting, exactly?

The National Domestic Violence Hotline defines gaslighting as a form of emotional abuse in which someone denies that events happened the way you know they did, so that you begin to question your memory, your judgment and your sense of reality.2 The signs it lists are mostly internal: second-guessing yourself constantly, asking “Am I too sensitive?” many times a day, apologizing all the time, and making excuses for the other person to friends and family.2

Here’s the part most people find surprising. Sociologist Paige Sweet, interviewing 43 women who had survived domestic violence, argued that gaslighting is not only a psychological trick. It works best when it borrows from things the wider world already believes, such as stereotypes about who is “emotional” or “unreliable”, so that the person being doubted is easier to doubt.3

A 2023 study by Willis Klein, Sherry Li and Suzanne Wood collected accounts from 65 people, aged 18 to 69, who said they had been gaslit by a partner.4 Two motives came up again and again: avoiding accountability for something (infidelity was a common example) and controlling what the other person did.4 The behavior rarely appeared alone. It traveled with intense early affection, isolation from friends and family, unpredictable swings between warmth and coldness, and cold-shouldering.4

Gaslighting has also moved from a pop word to something researchers measure. A recent measure, the Gaslighting Relationship Exposure Inventory, asks 11 questions about a romantic partner’s behavior and was tested with 509 people in Israel and 395 in the United States who were in relationships.5 That matters because it lets researchers study how often this happens outside of crisis settings, not only among people who already identify as survivors.

Translated into what you can actually see: the question is not whether two people remember Saturday differently. Memory is unreliable for everyone. It is whether one person’s version always wins, and whether you are paying for disagreeing.

Where do silence and control fit in?

Researchers who measure emotional abuse do not treat it as one thing. In a 1999 study of 157 female undergraduates in dating relationships, Murphy and Hoover found support for four distinct forms: restrictive engulfment (isolating and monitoring a partner), hostile withdrawal (punishing with coldness and refusing to engage), denigration (humiliating and degrading) and dominance/intimidation (threats and aggression).6 Denigration and dominance were more closely linked to physical aggression than the other two.6

Notice that hostile withdrawal is on that list. Everyone needs a cooling-off hour sometimes. What Murphy and Hoover were measuring is coldness used as a penalty, the kind that ends when you concede rather than when the two of you talk.

When several of these forms combine and your choices keep narrowing, advocates use the term coercive control: a pattern, not an incident. That is the point at which this stops being a communication problem and becomes a safety question.

How is this different from an ordinary fight?

Ordinary conflict can be loud, unfair and painful. What tends to separate it from manipulation is what happens afterward. After a fight, both versions of events can survive, and the person who was wrong can say so. In a manipulative pattern, one version wins every time, and the cost of holding yours keeps rising.

Here is Rania’s Friday, pulled apart:

• What happened: Gideon texted “sure” about the dinner, later said he had not agreed, and replied in single words for two days until she apologized.
• The reading: “He’s rewriting what happened so he doesn’t have to come.” That is a reading, and it might be wrong. He may honestly have meant “sure, maybe.”
• The fear: “Maybe I really am forgetful. Maybe I’m the problem.”
• The next step: Notice whether this is the first time or the fifth, and say what she remembers plainly, once, without needing him to agree.

One Friday tells her very little. A season of Fridays like it tells her a lot.

What to do this week

• Keep your own record. A private note with dates and what was said, not to win an argument but to steady your own memory.
• Test one small disagreement. Hold a minor position calmly (“I remember it differently, and that’s okay”) and watch what it costs.
• Tell one person the unedited version. If you have been softening the story for friends, that softening is information.
• Watch the direction of travel. Are you seeing fewer people, explaining more plans, apologizing more often than a few months ago?
• Get an outside view. A licensed counselor can help you sort this. If you feel afraid of the other person, or your freedom is shrinking, an advocate at a domestic violence hotline can talk it through confidentially, without you having to name it first.

What doesn’t the research show?

• The manipulation tactics study described everyday influence, not abuse, and it did not set a line where one becomes the other.1
• The gaslighting interviews and surveys were with people who identified as survivors. They show what the experience looks like from inside, not how common it is.34
• A questionnaire is a research tool, not a self-test. The gaslighting inventory was built to study groups of people, and a score on it would not tell you what is happening in one relationship.5
• The four-form measure was developed with female undergraduates in dating relationships, so it may not map perfectly onto every age, gender or relationship type.6
• No checklist can tell you what another person intends. The research describes patterns and their effects on the person on the receiving end, which is exactly why your own record and an outside view matter.