The idea that the loudest person in the room is compensating for something is one of those claims that feels too obvious to check. It is also, unusually for folk wisdom, a real hypothesis that psychologists have spent about thirty years trying to test, and the results are more interesting than either the believers or the debunkers tend to report.
It has a name. Researchers call it the mask model.
What the mask model actually proposes
The idea comes out of psychodynamic theory: that outward grandiosity is a defensive structure built over the top of self-doubt. The confident surface is not the person. It is the thing protecting the person.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
In a 2018 chapter in the Handbook of Trait Narcissism, Sophie Kuchynka and Jennifer Bosson set out the three specific forms this claim can take, which is where the folk version starts to come apart. It could mean high stated self-esteem sitting on top of low automatic self-esteem. It could mean high but unstable self-regard that swings with events. Or it could mean high self-regard that is conditional, holding up only as long as the person keeps succeeding.
Those are three different claims about three different things. Ordinary conversation treats them as one, and they do not stand or fall together.
Why the results kept coming out inconsistent
The first version, the hidden-low-self-esteem one, is the version people mean when they say someone is overcompensating. It is also the one with the messiest evidence.
Bosson and colleagues reviewed this directly in Untangling the Links between Narcissism and Self-esteem, published in Personality and Social Psychology Compass in 2008. Their summary is that some investigations found the masking pattern and others failed to replicate it, and much of the paper is spent on why.
Three reasons stand out in their account. The tests used to measure automatic, unstated self-esteem are not very reliable, which alone can produce a literature that appears to disagree with itself. Narcissism relates differently to self-views about competence and drive than to self-views about warmth and connection, so a study measuring one can contradict a study measuring the other without either being wrong. And the trait itself splits into grandiose and vulnerable forms that behave differently.
That third point deserves a moment, because it is the one most likely to be doing the damage in everyday conversation. Grandiose narcissism is the outward, self-assured, attention-taking version. Vulnerable narcissism is the version that is entitled and self-focused but also anxious, thin-skinned and prone to withdrawal. Both are measured in this literature. They are not the same people, and the person who fits the second description is generally not the loudest person in any room.
So a claim about loud people has, at best, been tested on one of the two groups it is usually taken to describe.
A pile of studies that disagree is not the same as a finding. It is usually a sign that the question was asked with instruments that were not up to it.
The version that did hold up
Kuchynka and Bosson’s conclusion is the useful part. Of the three hypotheses, they judge that the contingent self-esteem version has the most support: the fragility is real, but it lives in how much a person’s self-worth depends on succeeding at things, particularly achievement.
That is a meaningfully different claim from secret self-hatred. Someone whose good opinion of themselves is conditional on winning does not necessarily hold a low opinion of themselves underneath. They hold a high one that has to be paid for continuously.
The two look similar from outside and behave differently. Hidden low self-esteem would predict someone who crumbles when praised, because the praise conflicts with what they believe. Contingent self-esteem predicts something closer to the opposite: someone who needs the praise, responds badly to its absence, and is hardest to be around not when they are being challenged but when they are losing.
The authors’ own recommendation is that researchers stop looking for masked self-hatred and start looking for fragility in achievement situations, which is a fairly clear signal about where the previous three decades went.
What none of this licenses
This is where the research tends to get stretched, including by people quoting it accurately.
None of these studies measured loudness. They measured a trait using questionnaires, and then measured self-esteem using other instruments. Volume in a meeting was never the variable. The chain from “this person talks a lot” to “this person scores high on trait narcissism” to “this person has contingent self-worth” has two weak links in it, and the research addresses only the second one.
Plenty of people are loud because they are enthusiastic, or hard of hearing, or grew up in a big family, or are the only person in the room who finds the topic interesting.
It is also worth noticing what the saying does for the person repeating it. It converts someone who is taking up more space than you would like into someone secretly worse off than you, which is a satisfying trade and requires no evidence. That is an observation about how the phrase gets used, not a research finding, but it may explain why a claim this shaky has survived this long.
The other thing worth saying plainly: narcissism in this literature is a personality dimension measured on a scale, not a diagnosis, and nothing here is a tool for working out what is wrong with a colleague. Describing what someone does is one thing. Deciding what they secretly feel is a different activity, and this research is better read as a caution against it than a license for it.
What we keep coming back to is how much more specific the supported version is than the saying. The evidence does not say confident people are secretly fragile. It says that for some people, high self-regard is expensive to maintain, and that the cost shows up where achievement is on the line rather than in the volume of their voice.