There is something happening at work. It has been happening for a while. A colleague takes credit for things they did not do, consistently and without apparent embarrassment. A manager dismisses your contributions in meetings full of people, then acts on them privately a week later. A process is visibly failing, everyone in the team can see it, and nobody says so in the room where it could actually be fixed. You sit with this. You mention it to people you trust, in careful, hedged terms, away from the people involved. You compose the conversation you would have in the shower, where it always goes well. And then Monday comes around, and you say nothing.
I see this constantly, across organisations, across seniority levels, across industries that consider themselves progressive and self-aware. What strikes me is not that it happens but how universal the mechanism is. The most competent people in a room will absorb the most obvious dysfunction quietly, for the longest time, and frame their silence to themselves as patience or professionalism or waiting for the right moment. The right moment, as a rule, does not come.
And the context in which all of this is happening has shifted. The years since 2020 have produced a labour market that lurches between tight and anxious with very little predictability. Redundancy cycles that once happened quietly and occasionally now happen publicly and repeatedly. Entire industries restructured around AI before anyone has quite worked out what that means for the people inside them. Cost of living pressures that make the prospect of losing a job not just professionally inconvenient but genuinely frightening. In that environment, the calculus around speaking up has changed. The stakes feel higher because, in a number of real ways, they are.
What the brain is actually doing
When we perceive a social threat, including being judged, excluded, or losing status, the brain activates its threat response in a way that is neurologically similar to a physical danger signal. Research by Naomi Eisenberger at UCLA demonstrated that social pain, the pain of rejection, exclusion, or humiliation, activates many of the same neural regions as physical pain. The brain is not overreacting when it treats a difficult workplace conversation as dangerous. It is treating it as dangerous because, to the structures responsible for self-preservation, that is precisely what it is.
There is also what psychologists call motivated reasoning at work. The brain is not a neutral processor of information. It is motivated to reach conclusions that preserve comfort and avoid threat. Which means that when you are sitting with something that needs to be said, your brain is simultaneously constructing a very convincing case for why now is not the right moment, why the person probably did not mean it that way, why raising it will only make things worse. These arguments feel like logic. They are, in significant part, your threat response making itself sound reasonable.
"The brain does not distinguish clearly between social threat and physical threat. Your silence is not cowardice. It is your nervous system doing exactly what it was built to do. The problem is that it was built for a different set of threats."
Layer on top of this the bystander effect, the well-documented phenomenon in which people are less likely to act in a situation the more others are present. In group settings, responsibility diffuses. Everyone assumes someone else will raise it. Everyone is waiting for evidence that it is safe to go first. Nobody goes first. The meeting ends. The problem persists. I have watched this happen in rooms where every single person, individually, knew exactly what the problem was.
What is notable about the current moment is that economic anxiety amplifies all of these mechanisms. When job security feels precarious, the social threat of speaking up becomes entangled with a more concrete threat: what if this costs me my position? Research on workplace silence shows that employees are significantly less likely to raise concerns during periods of organisational uncertainty, not because the concerns disappear but because the perceived risk of raising them rises. The silence is not complacency. It is a rational adaptation to a more threatening environment. The fact that it is rational does not make it sustainable.
What you actually risk if you do say something
Most advice skips straight past this part, which is part of why the advice rarely lands. The risks are real and they are not evenly distributed.
Research on workplace retaliation consistently shows that people who raise concerns, even through formal channels designed for the purpose, face social penalties. They are perceived as less committed, less of a team player, more likely to cause problems in future. This perception can be invisible and deniable, never showing up in a performance review but absolutely showing up in who gets included, who gets the interesting work, who gets considered when something good becomes available. Women and people from minority groups face a steeper version of this gradient. The person who names the problem frequently becomes the problem, at least in the short term.
In a market where headcount is being watched, where restructuring is a constant background noise, where being seen as low-maintenance is not a neutral quality but an active asset, that social penalty carries more weight than it did five years ago. I am not making the argument for silence. I am making the argument for being precise about what you are walking into, because precision is the beginning of managing it.
The real risks, named plainly. What you can control: when and where the conversation happens; whether you frame it as a problem or an observation; whether you name behaviour or attribute motive; who else knows you raised it, and how; what you ask for, specifically, at the end of it. What you cannot control: how the other person receives it; whether the organisation responds well or poorly; informal social consequences that are never documented; whether the dynamic changes or simply moves underground; the long-term cost of continuing to say nothing.
How to de-risk it without making it a drama
The phrase "raising a concern" has been so thoroughly colonised by HR process that it now carries the weight of a formal complaint before a word has been said. The moment you frame something as a concern you are, in most people's minds, already filling in a form. The first thing to do is get out from under that framing entirely.
What the research on difficult conversations consistently shows is that the conversations that land well are specific, early, and framed around observable events rather than interpreted motives. "In the meeting on Tuesday, when you summarised my proposal, you did not mention I had written it" is specific and observable. "You always take credit for other people's work" is a character assessment and it will be received as an attack. One invites a conversation. The other invites a defence.
Timing and venue matter in ways that are chronically underestimated. A conversation that happens close to the incident, while it is still fresh and before resentment has built a supporting narrative around it, is structurally easier than one that happens weeks later, when both parties have had time to construct positions. The longer something sits, the more it grows. The more it grows, the higher the stakes feel. The higher the stakes feel, the longer it sits. This is the loop most people are stuck in, and the only exit is a specific, proximate conversation rather than a perfectly timed one.
Practical reframes - same issue, different entry points. Instead of "raising a concern": ask a question that requires them to engage with the specific moment. "Can I check something with you about Tuesday's meeting?" opens a door without announcing what is behind it. Instead of naming the pattern: name the instance. Patterns feel like accusation. A single specific moment feels like a conversation. Get the conversation first. The pattern becomes visible from inside it. Instead of explaining how it made you feel: explain what you need going forward. Feelings invite sympathy or argument. A clear, forward-looking request is harder to dismiss and easier to act on. Instead of waiting for the right moment: decide on a deadline. Not an ideal moment, a deadline. The ideal moment does not come. The deadline forces a specific, proximate decision rather than permanent deferral.
How to do this without looking like a victim or sounding an alarm
The fear of being seen as a victim is, for many people, as paralysing as the fear of retaliation. There is something about naming mistreatment that feels, in advance, like it will define you, like it will become the thing people associate with you from that point forward. This fear is not irrational. Tone is load-bearing in these conversations, and tone is the one thing most advice fails to address concretely.
The tone that works is curious rather than injured, direct rather than accusatory, and not hedged into incoherence. "I just wanted to check, and obviously I might have missed something, and I know it is probably not a big deal, but I was a bit surprised when..." arrives as an apology looking for reassurance. It does not arrive as a person who expects to be taken seriously. The hedge is there to manage the speaker's anxiety. It does not manage anything else.
Stating your own professional interest, quietly and plainly, is different from sounding an alarm. "I wanted to mention it because attribution matters for how my work is understood across the team" is not personal grievance. It is a person who knows the value of what they contribute, speaking about it as a professional matter. That version of the conversation does not require an audience, does not need to go anywhere beyond the room, and does not make the person delivering it look fragile. It makes them look like someone who pays attention and expects others to do the same.
The cost of doing nothing, and why it is rising
Research on emotional suppression shows that the cognitive effort involved in managing an unaddressed grievance, rehearsing what you would say, monitoring interactions for further incidents, maintaining a surface of normality while processing something underneath it, is genuinely costly. It consumes working memory. It affects concentration. Over time it corrodes the baseline sense that work is a place where fair outcomes are possible, which turns out to matter quite a lot to whether you can function well in it.
What I observe, and what the research supports, is that this cost is not evenly distributed across time. The longer you wait, the more the unspoken thing accumulates. Other incidents attach to it. A story forms. By the time you finally feel compelled to say something, it is no longer one conversation about one moment. It is a case, and cases feel like confrontations, and confrontations feel like nuclear options, and so most people never deploy them and instead quietly start updating their CV.
In the current climate, that quiet exit is increasingly expensive too. Mobility has contracted. The next job is not always there, or not quickly, or not on equivalent terms. People are staying in situations they would previously have left, which means the unspoken thing has more time to grow, and the pressure inside the organisation builds in ways that tend not to end well for anyone.
The conversation you have been composing in the shower does not have to be a reckoning. It just has to be a conversation: specific enough to be answerable, calm enough to be heard, and early enough that it is still about one thing rather than everything. That is a considerably lower bar than the one your brain has been constructing. Your brain, as noted, has its own reasons to keep raising it. You do not have to let it.
Kate Southerby writes on neuroscience, consumer behaviour, and the psychology of decision-making. References: Eisenberger (2012) The pain of social disconnection, Trends in Cognitive Sciences; Cortina & Magley (2003) Raising voice, risking retaliation, Journal of Occupational Health Psychology; Darley & Latane (1968) bystander intervention, Journal of Personality and Social Psychology; CIPD Workplace Conflict Report 2023; Stone, Patton & Heen (1999) Difficult Conversations, Penguin.
