Week 8: Who Has to Apologise?

Two colleagues, two mistakes, and only one of them produces a fast, personal apology — the difference is who can afford to let it sit

12 April 2027

Slides: pending upload

Two colleagues, same team, different rank.

A junior employee raises their voice once in a meeting, after weeks of pressure, and apologises within the hour — unprompted, specific, in person, to the senior colleague on the receiving end.

Weeks later, the senior colleague presents the junior employee’s report to leadership without naming who wrote it. The junior employee raises it. The reply arrives by email, after a delay, from someone else on the team: “there seems to have been some miscommunication about report authorship, and we appreciate you raising it.” No one apologises in the first person. No one is named as having done anything.

The second harm is arguably the more costly one — it affects the junior employee’s career, not just an uncomfortable ten minutes in a meeting. Only the first produced a fast, personal, first-person apology.

What actually differed#

Not the severity of the harm. What differed is who could afford to let the interaction stay unresolved. The junior employee has more riding on a smooth working relationship with someone senior, so a quick, specific apology partly repairs the harm and partly buys back that smoothness cheaply, before the friction can cost them anything. The senior colleague, protected by rank, can let the reply come from someone else, phrase it to remove any actor, let time pass before responding at all, or never personally respond — not because the mistake was smaller, but because the cost of leaving it unresolved falls somewhere else. This is Week 7’s control over timing and channel, read now as a power signal rather than a neutral variable: who gets to decide when and how a response happens is exactly the kind of control rank confers.

The institutional layer does the same work at a different scale. “There seems to have been some miscommunication” is Week 4’s agentless construction — no one to attach responsibility to — but inside a power relationship, the missing agent isn’t only a rhetorical accident. An institution can spend its apology on protecting the specific person whose decision caused the harm, by never letting a name attach to it in public.

What this isn’t claiming#

It isn’t that powerful people never apologise — sometimes a prompt, personal apology is cheap reputational insurance for someone who can easily absorb the cost of admitting fault, and plenty of senior people offer one for exactly that reason. And it isn’t that every pressure on a less powerful person to apologise is illegitimate — the junior employee in the first case may simply have been in the wrong, rank aside. What power changes is the conditions around the apology: who can afford delay, who controls the channel, whose mistake gets treated as requiring smoothing over versus absorbed as ordinary friction. It doesn’t mechanically decide who apologises and who doesn’t.

When rank and leverage point different ways#

Rank is a proxy for who can afford to let harm sit, not the thing itself. A specialist contractor whose departure would stall a launch can miss a deadline, apologise slowly and casually, and face no real consequence, despite having no formal authority over anyone on the team — the company’s dependence on their specific skills does the same protective work rank does for the senior colleague above. Meanwhile a senior manager whose position is publicly shaky after a bad quarter may apologise fast and personally for a comparable mistake, because their formal rank no longer buys them the safety it would in a stable position. What predicts the speed and specificity of an apology isn’t the title on either person’s contract; it’s who actually stands to lose more from the relationship souring — which rank usually tracks, but not always.

Today’s question was who’s obligated to apologise. Week 9 flips the lens: who has the power to decide whether an apology, once given, actually counts.