Article
The Difference Between Hard and Harmful
Not Everything is Toxic: Rethinking Conflict and Difference at Work. Part One of Three.
Workplaces are becoming less tolerant of ordinary disagreement and more likely to describe conflict in the language of harm. This three-part series explores how that shift happened, why it matters, and what leaders and organizations can do about it.
Our vocabulary for describing workplace difficulty has shifted. A tough manager is now described as toxic. A colleague who disagrees with our version of events is seen as gaslighting us. A co-worker we find difficult is recast as a narcissist, or someone we need boundaries from. The words sound clinical and feel accurate to the person using them. Often, they are neither.
Organizations have spent decades trying to address harassment and disrespect. That was the right thing to do. Real harassment causes real harm, and workplaces are better for having learned to recognize it. None of what follows is a call to turn back that progress. But every correction carries the risk of overcorrection, and that is the moment we seem to find ourselves in now.
Why does this matter? Because words shape how we understand problems and how we respond to them. Labels that were meant to identify serious misconduct begin to crowd out the language we need for the ordinary challenges of working with others. When that happens, disagreements lead to complaints and the possibility of recovery is foreclosed before anyone has really tried.
This article examines several labels that have become commonplace in workplace conversations, how the law has attempted to distinguish between what is hard and what is harmful, and why preserving that distinction matters.
A language that keeps expanding
Nick Haslam, a psychology professor, coined the term concept creep to describe the process by which the words to describe harm gradually expand to cover milder and broader experiences. In a workplace context, Haslam observes that what used to be called office politics, or simply a difficult boss, now arrives wrapped in far more serious language. [1]
Haslam emphasizes that this expansion is well intended and reflects a society more attuned to suffering than it once was, and that much of this is beneficial. But he notes that this benefit comes with a downside: everyday experience is pathologized.
The labelling of difficult behaviour
Several words now appear regularly in workplace conversations, often attributing far more weight than the situations they are being used to describe warrant.
The word “toxic” has become ubiquitous. What began as a way to describe genuinely corrosive environments now seems to attach to almost any workplace in which someone is unhappy. Frustration and disappointment are normal byproducts of working with others. The problem is that we have started to treat those feelings as proof of the diagnosis. If it feels toxic, the reasoning goes, then it is.
If “toxic” describes an environment, “boundaries” has become the term we use to manage each other within it. The concept itself is sound and genuinely useful. Boundaries help people protect their time, their wellbeing and their sense of what is acceptable. But the word has expanded beyond its intended purpose. The article “When Boundaries Are Weaponized” [2] describes a culture in which the language of limits is increasingly used to control others, or avoid a difficult conversation, rather than to communicate a need. A boundary in its healthy form is a statement about what a person will do. In its distorted form it becomes a demand about what everyone else must do, and a way to shut down dialogue rather than open it.
Part of what makes the word “boundary” so hard to question is that it presents itself as self-care. It is framed as protecting one’s mental health, and sometimes it genuinely is. But the same language can mask avoidance, giving the decision not to deal with someone the appearance of psychological necessity.
“Gaslighting” is another term that has drifted beyond its original meaning. It once described a sustained effort to make another person doubt their grip on reality. As the term became commonly used, mental health experts began to warn that it was being applied too broadly, to almost any disagreement about what happened and well beyond the deliberate manipulation it originally named. Someone who remembers a meeting differently, or honestly disputes another person’s account, is not gaslighting. But that term is now used in exactly those moments, carrying an unwarranted accusation of deliberate psychological abuse. [3]
Another clinical term finding its way into conversations about workplaces is “narcissist.” The CBC recently published an article entitled, “Narcissistic leaders more likely to oppose remote work, new research suggests.” The article reported on research suggesting that narcissistic dislike remote work because in-person work offers “richer channels for controlling and commanding reverence from employees.” [4] The research is not the concern. What is likely to be done with it is.
Genuine clinical narcissism is rare: the American Psychiatric Association estimates narcissistic personality disorder affects roughly 1-2 % of the population. [5] Yet the label is applied freely. A study like this one, however carefully done, can become a scientific-sounding justification for calling any leader who asks staff to return to the office a narcissist. A manager can make a decision an employee dislikes, and similarly, a colleague can be self-absorbed, without any personality disorder being involved.
When perception is considered proof
Some harms are concrete. For example, workplace violence can sometimes be easier to identify because there is usually a clear act: a shove, a thrown object, or a threat of physical harm. It happened or it did not.
Harassment requires a more careful assessment. Not because the experience of the person raising the concern is unimportant, or because harm must be visible to be real, but because how something is experienced does not, on its own, determine what it was. A perception may be sincere, and still need to be tested against context, conduct, and reasonableness. The task is to take the experience seriously while still asking what happened, what it meant in context, and whether it crosses the line from difficult to harmful.
How the law has tried to draw the line
When Ontario first added workplace harassment to the Occupational Health and Safety Act (OHSA) in 2010, it borrowed the definition of harassment almost word for word from the Human Rights Code. For six years, that broad definition stood on its own. Only with the 2016 amendments did the legislature add the crucial qualifier, now in section 1(4), that “A reasonable action taken by an employer or supervisor relating to the management and direction of workers or the workplace is not workplace harassment.”
It is hard not to read that as a correction, when it became evident that the unqualified definition was being used to attack legitimate employee management.
The statutory test reveals why harassment is so easily alleged and so often misunderstood. The OHSA defines workplace harassment as “a course of vexatious comment or conduct that is known or ought reasonably to be known to be unwelcome.” Intent plays no part. That is an important protection, but it also means the definition leans heavily on how the conduct was received. And that is when everyday conflict gets recast as a violation.
What vexatious actually means
Identifying harassment often turns on the word “vexatious.” In ordinary usage it simply means annoying, distressing, or agitating. Read that way, the definition would sweep in almost everything that irritates us at work, like the colleague who keeps arguing about a settled decision or the supervisor whose feedback stings.
“Vexatious” causes less trouble when harassment is tied to a protected ground under the Human Rights Code, such as sex, race or disability. There, the connection to the protected ground does much of the limiting work, because an employee must show not only that the conduct was vexatious but also that it occurred because of their membership in a protected group. Workplace harassment under the OHSA has no such limit.
That is why the test of harassment mandates some measure of objectivity. As Arbitrator Laing cautioned in an oft-quoted observation, harassment is “a serious word, to be used seriously and applied vigorously when the occasion warrants its use.” It should not be “trivialized, cheapened or devalued by using it as a loose label to cover petty acts or foolish words, where the harm, by any objective standard, is fleeting.” [6] In other words, an employee may feel offended, annoyed, or even intimidated by particular conduct, but those feelings alone do not determine the outcome. The question is whether the conduct was objectively unreasonable.
Conflict between colleagues
Two people can differ in countless ways that have nothing to do with mistreatment. One prefers to hash things out in the moment while the other needs time to think. One writes blunt, economical emails that the other reads as cold. One debates ideas openly in meetings while the other experiences that directness as an attack. None of these is a character flaw, and none is harassment. They are the normal result of different people, with different temperaments, being asked to cooperate closely.
The trouble begins when ordinary differences are recast as evidence of something more serious. The direct colleague becomes “aggressive.” The reserved one becomes “passive-aggressive.” A person who is difficult, blunt, needy, defensive, or whose personality clashes with ours becomes “toxic.” The focus shifts from the interaction to the identity of the other person. So rather than saying a person’s behaviour is toxic, we say the person is toxic.
As noted in an arbitration decision involving Cara Operations Ltd.: “There is a wide range of personalities that we experience in our interaction with others; not all of which may be pleasing to our individual sensitivities, but which we must live with nonetheless, within legal bounds.” [7] That should not be mistaken as a license for disrespect. It is a recognition that a degree of tolerance in the face of difference is part of what it means to work with other people.
Why the labels matter
This is not an argument for tolerating mistreatment. Harassment exists. Genuinely toxic environments exist. Narcissistic leaders and colleagues exist, and when they do, the impact is serious. The argument is narrower, and more practical. When these words are applied to ordinary conflict and human frailty, several things happen, all of which make matters worse.
The first is a loss of agency. The more we view ourselves as victims of harm rather than as participants in an unpleasant interaction, the harder it becomes to address the issue. A problem to be solved becomes a wound to be nursed. That leaves people less able, not more, to navigate the challenges inherent in working with others.
The second is dilution. When everything unpleasant is described as harassment or abuse, the words lose their meaning, in the way that awesome no longer conveys awe. Haslam makes the point precisely. As a concept expands, its centre of gravity shifts, so that the word calls to mind the lesser example rather than the grave one. The cost is borne by the people who have suffered the grave version, whose experience is inadvertently flattened into the same category as a curt email.
The third is institutional. Haslam warns that concept creep can unleash a flood of unwarranted accusations and invite enforcement that is disproportionate to the conduct in question. Anyone who has seen a minor disagreement escalate into a formal complaint, an investigation, and months of stress will recognize the pattern.
The fourth consequence is that a label ends the conversation. Once a colleague has been labelled toxic, a manager called a narcissist, or a disagreement reframed as gaslighting, there is nothing left to work out, because you do not negotiate with a diagnosis. The other person is no longer someone you are in a difficult moment with. They are only what the label says they are, and the only reasonable response to a toxic person or a narcissist is distance. In this way the language does more than exaggerate. It removes the possibility of repairing the relationship before anyone has tried.
A related move has the same effect. Over the years I have heard many people refuse to engage with someone they believe has harmed them until they have first received an apology. They treat an admission of fault as the entry fee for any conversation. It feels like accountability or even safety, but it operates like a locked gate. The other person cannot explain, cannot offer their side, cannot even establish what happened, because dialogue is withheld until they have already conceded they were wrong. A conversation that might have identified a misunderstanding never takes place, and the demand for an apology becomes a way to avoid the exchange altogether.
What we are witnessing is not confined to workplaces. We are seeing this between friends, who part ways over a disappointment or misunderstanding. We are seeing it in families, where a growing number of people are estranged from a parent, a sibling, or an adult child, often over the sort of rupture that earlier generations would have found a way to resolve or move past. This raises an obvious question: how did we become so quick to see difference, disagreement, and disappointment through the lens of harm? And why are we walking away rather than working towards a resolution?
What connects all of these, at work and beyond, is a gradual erosion of our capacity to remain in relationship with people we find difficult. Conflict is not a malfunction of human relationships. It is a normal, even necessary part of them. It is what happens when different people try to share a life, a family, a workplace, or a community. Learning to move through challenges without needing the other person to be a villain is one of the central skills of a shared life. The worry is that we are slowly losing the capacity to work through difficulty with other people and mistaking that loss for emotional wellbeing.
The next article explores how we got here, and why disagreement now feels so much harder than it once did.
[1]Nick Haslam, “Concept Creep: Psychology’s Expanding Concepts of Harm and Pathology,” Psychological Inquiry 27, no. 1 (2016): 1–17.
[2] Rachel Allyn, “When Boundaries Are Weaponized,” Psychology Today, December 8, 2025; see also commentary on cutting-off culture and the overuse of boundary language.
[3] Gaslighting,” Wikipedia, which collects the history of the term and reports concerns (including commentary in the Washington Post) that it has become a buzzword for ordinary disagreement.
[4] Sarah Petz, “Narcissistic leaders more likely to oppose remote work, new research suggests,” CBC News, July 18, 2026, reporting on research from the Wharton School, University of Pennsylvania.
[5] What Percentage of the Population Are Narcissists?”, ScienceInsights, May 4, 2026.
[6] British Columbia and B.C.G.E.U., Re, 1995 CanLII 18346 (BC LA).
[7] Cara Operations Ltd. and Teamsters Chemical, Energy and Allied Workers’ Union, Loc. 647 (Palmieri) (Re), 2005 CanLII 94066 (ON LA), at para. 24.
Lauren Bernardi is a lawyer and HR advisor. She is the founder of Bernardi Human Resource Law LLP and The Bernardi Centre, which help organizations navigate workplace issues and create psychologically safe, respectful, and inclusive workplaces.


