A woman opens our first call by describing a time her partner referenced something she’d never told him about. A restaurant she went to once. Or the name of a coworker she’d mentioned in passing, six months earlier. She sits with the moment, scrolls past, comes back to it later in bed. By the time she’s saying it out loud to me, she’s already braced for me to tell her she’s imagining it.
I won’t. The behavior has a name in the research literature, though the popular shorthand “covert narcissism” is doing more work than it should. Researchers more often use the term vulnerable narcissism, which lives in academic literature rather than the DSM. What follows applies to a behavior cluster I see in my practice. Not a clean clinical category.
What’s been on my mind across the last few years is the surveillance, and the way it isn’t constant. It thickens around certain weeks of a relationship and recedes during others. After eight years of consulting work I can describe four of the moments when it sharpens and you can feel what covert narcissist surveillance looks like up close.
TL;DR
Vulnerable Narcissism Drives It: Covert narcissist surveillance maps to the research construct of vulnerable narcissism; Aaron Pincus’s group at Penn State built a dedicated surveillance-use scale tied to it.
Attention Is Inventory: Early questions about your finances, old roommates, and support network operate as data collection; the more he knows, the less exposed he feels.
Independence Triggers Monitoring: Paul Wink’s 1991 hypersensitivity work and recent social-platform research show imagined withdrawal lands the same as actual withdrawal; your better mood escalates surveillance.
CBT Boundaries Backfire: In my caseload, standard boundary advice treats the surveiller as anxious instead of coercive, hands him a counter-script, and leaves you apologizing for drawing the line.
Deniability Is Structural: Every surveillance behavior arrives pre-explained (“safety alert,” “stranger account,” “emergency tracking”), so no single incident reaches a third party as evidence.
Before any of it looked like surveillance, it looked like attention
Zelda came to me in fall 2022. She was a pediatric ICU nurse studying for her CCRN exam, and her stack of study cards was visible on the bookshelf behind her on the video call. Her partner of three years had remembered, six months in, the name of a substitute teacher she’d mentioned once on their second date. She told me about the substitute teacher in our first session because it was the first thing that had felt, retroactively, off.
I taught this stage of the behavior wrong for the first two years of my practice. I taught it as jealousy. Zelda’s case changed that. She told me about the night her partner showed up at her hospital cafeteria at 11 PM with takeout, after she’d told him she was studying late. He framed it as a surprise. She thanked him in the moment. Six months later she walked it back. She’d never told him which break room she’d been using, and he’d come straight to it.
The early attention had been functioning as inventory. Questions about a client’s mother. Curiosity about old roommates. Aaron Pincus’s group at Penn State has a surveillance-use scale tied to vulnerable narcissism, and the dynamic Zelda described matched it closely. The more her partner knew about her, the less exposed he felt.
Real attention exists too. I’m watching for direction. If the questions cluster around her finances and the people she’d turn to if things got hard, that’s a different shape than someone learning who she is.
Session four
Klarity does medical billing from her kitchen in a small town outside Cleveland. Fall 2021. A kettle whistled in our second session and she let it whistle for a minute before she got up. She’d been crying before the call. She said “I think I’m imagining it” four times in the first hour. I noted it and didn’t push back yet.
The week before our fourth session she’d finished a billing project that paid well, and taken herself out to dinner with a friend. She went to bed feeling, in her words, lighter than she had in months. The next morning her partner asked her, casually, whether the restaurant was new. She’d never told him about the restaurant. He’d assembled it from a credit card alert he’d asked her to set up “for safety” eighteen months earlier and never turned off. If she’d told her sister, the sister would have heard a thoughtful man. He set up a safety alert and never turned it off. Every word of that is true.
Most survivors misread this moment. They assume the monitoring sharpened because their partner noticed something specific. From my caseload, what flips the switch is subtler. Recent work on vulnerable narcissism and surveillance behavior on social platforms finds that imagined withdrawal lands the same as actual withdrawal in this group, and Paul Wink’s 1991 work on covert-narcissistic hypersensitivity describes the same mechanism from a different angle. Her partner experienced her good mood as if she’d already left.
I started taking these consultations at Embrace Inner Chaos in 2018, and for Klarity I made the wrong call. I gave her the standard advice. Communicate. Set a clear boundary. She came back two weeks later more confused than before because her partner had a ready response to every boundary that left her apologizing for having drawn it. I’d handed her language he used against her, and she spent a fortnight defending herself with it before she gave up and called me back. The cost wasn’t mine. It was hers.
The standard CBT-derived approach treats the surveiller as if he’s an anxious partner with reparable habits. He isn’t.
After Maude said no
Maude was an ER charge nurse in Phoenix. Summer 2022. She joined our call from her car in the hospital parking garage, still in scrubs, the engine off because she didn’t trust the air conditioning to last the session. She’d been a forensic nurse early in her career, so she knew how to document. Even she felt she couldn’t prove what she was describing.
She’d told her husband, the previous Sunday, that she wasn’t going to his family’s lake weekend. The first no she’d issued in two years. Nothing big happened. He said okay. He said he understood. Three days later she noticed the shared location app, set up for emergencies during a hospital stay the year before, was being opened at hours that didn’t match his schedule. The app logged it. She’d never agreed to leave it on.
Green and Charles documented this dynamic in their qualitative analysis of vulnerable narcissism, paranoia, and control. Small rejections register as catastrophic in this group, and the response often goes underground rather than escalating openly.
The day a survivor holds a line is the day a lot of my consultations get more urgent. Maude is the case where I started teaching, internally, that boundary work with this kind of partner is a different intervention than boundary work with a securely attached one. With Maude I shifted us toward documentation, away from communication. That isn’t standard couples-therapy advice, and a therapist who disagrees with me should push back on it.
When she’s gone
Willow is a paralegal at a small immigration firm. She came to me in August 2023. We did the call from her car because her teenage daughter was home and Willow didn’t want her overhearing. Six months out from the divorce. Fully no contact. Every account blocked. Then an Instagram comment showed up on a photo from 2019, from a handle she didn’t recognize. The comment said “remember the cabin.” One person knew about the cabin.
She turned the phone around so I could see the screenshot. She’d zoomed in on it. Her hand was steady, but her thumb was tapping the side of the case the way some people’s hands do when the rest of the body is trying not to move. She didn’t cry. The not-crying is the part I sat with after the session.
Before Willow I’d been seeing post-discard contact mostly through mutual friends, what popular writing calls “flying monkeys”. I think the term is overused. Sometimes a friend asking “are you doing okay?” is a friend asking, not a recruited intermediary, and the discernment is harder than the popular literature admits.
Willow taught me the layer beneath that. Throwaway emails and virtual numbers. Anonymous comments timed to her birthday, and one timed to her daughter’s first day of high school. That’s the one I come back to.
To time a comment to a fourteen-year-old’s first day of school is to have kept track of which school, which year, which week. Each comment carried its built-in deniability. None of them gave Willow anything she could put in a filing.
What I see in the body when a client describes receiving one of these is a specific contraction. Her shoulders drop a quarter inch, her voice flattens. The nervous system has already recognized the sender; the conscious mind hasn’t been given enough to act on. Research on social exclusion and increased surveillance behavior finds the surveilling partner’s behavior intensifies after the relationship ends, the opposite of what most survivors expect when they finally leave.
Azalea was a community college instructor. Winter 2020. She’d left a partner who, in her words, “hated my work and never came to my talks.” Six months after the breakup he turned up at an academic event she presented at, four states from where they’d both lived. He told a mutual friend he happened to be in town for unrelated work. The friend believed him. Azalea saw him in the audience during her Q&A. When she first described this to me, I assumed it was a coincidence and a story her trauma was attaching to. I told her so in the first session. She came back. I was wrong, and she’d come to me to be believed, and the first thing I did was doubt her. The event dates had been on the college’s site for three months.
There’s a different case I think about more. Nevaeh teaches third grade in central Pennsylvania. She came to me in 2021 already convinced her partner was monitoring her phone. We were two sessions in and I was nudging her toward documentation, the Maude playbook. She came to session three and told me her partner’s mother had been dying since March. He hadn’t told her until the previous weekend. The withdrawal Nevaeh had been reading as monitoring was him sitting in hospital parking lots for three months, not knowing how to say it. I’d been six minutes from telling her to start screenshotting his texts.
Nevaeh stopped coming after the fifth session. I think I’d lost her by then. I don’t know what I should have done differently, and I’m tired of pretending I have a tidy answer for it. The base rate of genuine attentiveness in the general population is much higher than the base rate of surveillance, and a coach who only sees the worst cases will start to see the worst cases everywhere. I notice this in myself and try to correct for it.
At this stage, in clients who do leave, what I see most is grief. The slow recognition that the access doesn’t fully close, and that proving any single incident would take more effort than any court or family member is going to invest.
Why it stays unprovable
The behaviors above come bundled with their explanations. The shared location app was for emergencies. The credit card alert was for safety. The Instagram comment came from a stranger account. The 11 PM takeout was a surprise. Every behavior arrives with a sentence already attached that makes it sound reasonable to an outsider.
Sam Vaknin, whose work is popular online but operates outside formal clinical research, frames this directly in his analysis on privacy as the opposite of narcissism. The arrangement depends on dismantling another person’s privacy while protecting the narcissist’s own. I don’t take all of Vaknin’s framing on board (he reads partner behaviors as more uniform than my caseload does), but the core observation lines up.
The limits of what I can claim matter. Not every partner who remembers small things is taking inventory. Nevaeh is the case I keep close. I’d rather a client leave a consultation less certain about her partner than more certain.
When several of these behaviors cluster and each arrives pre-explained, what makes a reader’s stomach drop isn’t any single incident. It’s the layer of plausible reasonableness sitting on top of all of them at once. The phrases that maintain that layer are recognizable across cases. “You’re being paranoid.” “Why would I care enough to check.” Neither contradicts her perception directly. Each offers her a more comfortable explanation than the one her own eyes gave her, and the more times she hears it the more accessible that explanation becomes. Six months in, when she tries to retrieve what actually happened, the explanation comes up first. Her own observation comes up second, slower, with less confidence in it.
Some popular writing says the deniability is “the abuse itself.” Some claims hypervigilance in a survivor is always correct reading. Both are too clean. Hypervigilance is sometimes correct reading and sometimes a trauma response misfiring. I’ve sat with women whose suspicions were accurate. I’ve sat with women whose suspicions weren’t, and who were grieving a different wound. I get it wrong sometimes.
If you recognized your life somewhere above, you already know which sentence you stopped at. I’m not going to tell you what to do with it.
You don’t have to figure this out alone.
Work with Mr. Som Dutt — a coach who understands narcissistic dynamics inside out.
Whether you’re a survivor or recognizing narcissistic patterns in yourself — I help you see the cycles clearly and move forward.
For Survivors
Partners, spouses, parent/child, siblings, friends & colleagues — see the patterns, reclaim clarity.For Narcissists Seeking Change
Awareness is the hardest step. I help you understand and manage your behaviour.FAQs
What Is Covert Narcissist Surveillance?
Covert narcissist surveillance is hidden monitoring tied to the research construct of vulnerable narcissism, where partner observation operates through plausible explanations rather than direct confrontation. Aaron Pincus’s group at Penn State developed a dedicated surveillance-use scale measuring this pattern in clinical populations.
Why Does Covert Narcissist Surveillance Intensify When I’m Doing Better?
Research on vulnerable narcissism and Paul Wink’s 1991 hypersensitivity work show imagined withdrawal lands the same as actual withdrawal for these partners. Your good mood gets read as you already leaving, which escalates monitoring even when nothing about your behavior toward him has changed.
How Does Covert Narcissist Surveillance Differ From Genuine Attention?
Real attention is about learning who you are. Covert narcissist surveillance functions as inventory: questions cluster around your finances, old roommates, and support network, building a file the partner uses later rather than expressing curiosity about your inner life.
Why Can’t I Prove Covert Narcissist Surveillance To A Third Party?
Every surveillance behavior arrives pre-explained: the location app was for emergencies, the credit card alert was for safety, the late-night appearance was a surprise. Each story is true on its face, which is why no single incident reaches a friend, family member, or court as evidence.
Does Covert Narcissist Surveillance Stop After The Breakup?
No: research on social exclusion and surveillance behavior finds monitoring intensifies after the relationship ends, not fades. Throwaway emails, virtual numbers, and timed anonymous comments allow continued access while preserving total deniability.
Why Did Standard Boundary Advice Backfire In My Covert Narcissist Relationship?
Standard CBT-derived couples-therapy approaches treat the surveiller as an anxious partner with reparable habits, when he isn’t. Boundary-setting assumes shared reality between both people, and that assumption is the thing being exploited.
What Triggers Covert Narcissist Surveillance To Sharpen Suddenly?
Two moments most consistently sharpen surveillance in my caseload: when she gets her life slightly back and feels lighter, and when she issues her first clear “no” after a long period of compliance. Green and Charles’s qualitative analysis of vulnerable narcissism and control found small rejections register as catastrophic in this group.
Is Covert Narcissism A Clinical Diagnosis In The DSM?
The popular term “covert narcissism” doesn’t appear in the DSM. Researchers more often use vulnerable narcissism, which lives in academic literature as a personality construct rather than a clinical diagnosis.