Empathy, Weaponised.
Someone's distress can move us to help before we know what happened. The difficulty begins when concern for that person becomes certainty about another. Shared feeling can carry an accusation further than the evidence supporting it.

A friend tells you that someone has treated them cruelly. You can hear that they are upset, and you begin to feel angry on their behalf. By the time they finish, you may already dislike the person they are describing. You have learned something directly about your friend's distress. Your knowledge of the other person's conduct has arrived through an account of it. In the moment, that difference can be difficult to notice.
By empathy we may mean understanding another person's perspective, sharing something of their feeling, or both. Emotional contagion is related but narrower: we can take on a mood without clearly understanding its source or distinguishing it from our own. Compassion adds concern for the person and an inclination to help. These experiences can overlap, but treating them as one mechanism makes it harder to see where a moral judgement enters.
Understanding that someone feels betrayed is already useful. It may tell us that they need company, reassurance or practical assistance. Establishing whether a betrayal occurred, who was responsible and what response would be justified requires further information. The person's suffering is relevant to those questions without answering all of them.
The temptation is to let the emotion carry its explanation with it. I feel your anger, so the event begins to appear as you describe it. If you are a friend, doubting the account can also feel like failing you. I may reassure you by agreeing about somebody else's motives when what I can actually establish is that you were hurt. Support has quietly acquired a factual claim.
Now imagine the story passing through a group. Each person encounters both the original account and the reactions of people they trust. The next response feels more justified because others have responded similarly. Yet those reactions may all depend on the same source. Ten people can independently feel upset without providing ten independent reasons to believe the allegation.
The process can also change the story's emphasis. The detail that carries the strongest moral charge is repeated, while qualifications are omitted because they take longer to explain or seem to weaken the expression of support. A tentative account becomes easier to remember once it has a clear victim, a clear offender and an obvious response. Nobody needs to coordinate this change for it to become consequential.
Online communication gives the process speed and reach. Research on moral and political discussion has found associations between moral-emotional language and greater message diffusion. That does not mean every widely shared account is false, or that emotion alone explains its reach. It does suggest a reason to be careful about treating visibility as independent evidence of accuracy. The features that make an account travel are not necessarily the features that make it well supported.
Deliberate manipulation can use the same opening. Someone may present suffering selectively, conceal relevant context or insist that sympathy requires an immediate judgement against a target. The person responding may feel morally active while their attention has been arranged around only part of the situation. Genuine suffering can be used this way too. A story does not have to be fabricated for the conclusions drawn from it to exceed what it establishes.
There is an obvious difficulty with asking for more examination: people who have been harmed are often already struggling to be heard. An insistence on complete certainty can become a way of indefinitely withholding help. The distinction has to appear in the response itself. We can listen, offer practical support and take proportionate precautions while keeping unresolved claims open. Helping a distressed person and issuing a final judgement about someone else need not be the same act.
Nor does careful judgement require feeling equally for everyone at every moment. Affection, attention and proximity are uneven. A friend will usually matter to us in ways a stranger does not. The question is whether that partiality is permitted to decide what happened. I can care particularly about you while applying a standard of evidence I would also want used if somebody accused you.
For someone who finds another person's tension difficult to bear, agreement may bring immediate relief. If it calms the other person and reduces the discomfort in the room, repeating it can become easier than saying, “I understand why you are hurt, but I do not know whether that is what they intended.”
Sharing distress can become burdensome, especially when someone has little opportunity to recover or feels responsible for resolving every emotion they encounter. Caring for a person does not require matching the intensity of their suffering. The distinction between empathic distress and compassionate concern is useful here: becoming overwhelmed and remaining willing to help are different responses. Neither the amount of pain we absorb nor the exhaustion we accumulate measures the quality of our care.
The same distinction matters at the scale of a group. An atmosphere of anger can make hesitation appear cold, while agreement feels warm and immediate. Before adopting the group's certainty, it helps to ask which part we have encountered ourselves, which part is another person's account, and which part is an inference about motives. A pause can preserve the ability to help without adding an accusation we are not yet entitled to make.
Empathy becomes easier to weaponise when concern is made conditional on accepting an explanation. Refusing that condition may be difficult precisely because the feeling is sincere. We can take another person's suffering seriously and remain answerable for what we do with it. The responsibility begins where their account becomes our judgement.
Further reading: Decety and Jackson: the functional architecture of human empathy; Brady and colleagues: emotion and the diffusion of moralized content; Klimecki and colleagues: empathy and compassion training.
