Describe, Don’t Blame: Observation-Based Language
Making room, in the last chapter, gets more voices into the conversation, including the ones that name what is going wrong. And that is exactly where a lot of trust is won or lost. The moment you point at a problem, your words decide whether the other person leans in to help fix it or braces to defend themselves. This chapter is about saying the hard thing in a way that keeps the room safe: describing what happened, instead of blaming the person who did it. It is the next move to write into your project’s Inclusive Communication Plan, and one of the most trust-protecting of them all.
Tell an Observation from an Evaluation
Picture a group of volunteers who have just missed the deadline for a community newsletter. One member turns to another and says, “You’re so disorganized, that’s why we’re late.” Now imagine the same worry put a different way: “The draft came in two days after the date we’d agreed, so we missed the print slot.” Both point at the same problem. Only one of them starts a fight.
The gap between those two sentences is the gap between an evaluation and an observation. An observation is what a camera would have recorded: what was actually said or done, with no judgment attached. “The draft arrived two days after the date we agreed” is an observation, you could hold it up against the calendar and the email. An evaluation is a judgment about the person: “you’re disorganized,” “you’re careless,” “you don’t take this seriously.” No camera can record “disorganized.” It is a conclusion you have drawn about who someone is.
Leading with the observation is the opening move of Nonviolent Communication, an approach to difficult conversations developed by Marshall Rosenberg. You do not need the whole method here. The one piece worth taking from it is this: separate what happened from what you make of it, and say what happened first.
Here is why that small separation earns a whole chapter. An evaluation lands as an attack on the person, so the person defends themselves, and once someone is busy defending who they are, they have stopped listening to the actual problem. An observation lands differently: it is a fact set on the table that the two of you can look at together. Same concern, completely different conversation. One puts a person on trial; the other puts a problem in the middle of the room.
But what if the person really is being careless? Am I supposed to pretend I don’t think that?
No. You are allowed to think it. The point is not to hide your view, it is to lead with the observable thing that gave you that view, because that is the part the other person can actually do something about. “You’re careless” gives them nothing to fix but their character. “The last two entries were logged under the wrong date” gives them a specific, fixable thing, and leaves their character out of it.
Turn a Judgment into an Observation
When a judgment is on the tip of your tongue, you can convert it into an observation in three steps. It feels mechanical the first few times, and then it becomes automatic.
Catch the evaluation. Notice the judgment word forming: “lazy,” “rude,” “disorganized,” “always,” “never.” Those words are the tell that you are about to describe a person instead of a behavior.
Find the observable behavior behind it. Ask the camera question: what did I actually see or hear that led me to that word? Swap the judgment for the specific thing, at a specific time, that a camera would have caught.
State the behavior and its effect, not a verdict. Put the two together: what happened, and what it led to. “The minutes went out the night before the meeting” (behavior), “so a couple of us hadn’t read them in time” (effect). No conclusion about the person is needed; the behavior and its effect make the point on their own.
Why Does This Protect Trust?
Step back to Part 1 for a moment. Trust, we said, is your willingness to be a little vulnerable with people because you believe they will act well toward you. Psychological safety is the shared sense that speaking up will not be punished. Blame is the fastest way to burn through both at once. The moment a hard conversation turns into a verdict on someone’s character, everyone watching learns the same lesson: mistakes here get personal, so hide them. That is a low-safety team forming in real time.
Observation-based language is what lets a team say hard things without paying that price. Because you are describing a behavior and not indicting a person, the member on the receiving end can take the point without feeling attacked, and everyone else sees that a problem can be raised here without it turning into a fight. You get to keep the honesty and the trust at the same time, which, back in Part 1 Chapter 3, was the whole definition of a team that runs on high safety and high standards together.
Rephrase Three Blaming Lines
Let us watch the move on three real sentences. A community garden is run by a group of volunteers who share plots, tools, and a watering rota. Like any group, it hits friction. Here are three things a frustrated member might say, each turned from a blaming line into an observation and its effect.
“You never do your share of the watering.” That is an evaluation: “never” and “your share” are verdicts, and the line invites a flat denial. The observation underneath: “The watering log shows the Tuesday slot was missed the last three weeks, and some of the seedlings have dried out.” Now there is a specific, checkable fact and a real effect, with nothing to deny, only something to sort out.
“You’re territorial about the good tools.” “Territorial” is a character verdict. What did a camera see? “The new shears have been kept in your shed for the past month, so the rest of us have been using the blunt pair.” Same concern, but now it is about where the shears are, not about what kind of person you are.
“You just take over every meeting.” “Take over” is a judgment, and “every” guarantees an argument. The observable version: “In the last two meetings, the plot assignments were settled before the newer members had said anything.” That points at a pattern anyone can look at, and it opens the door to changing it rather than defending against it.
Notice the shared shape. Each rewrite drops the character word, names the specific behavior a camera could have caught, and adds the effect it had. None of them pretends there is no problem. They simply put the problem, rather than the person, in the middle of the table.
Your Turn!
Now you make the move yourself, on paper.
The situation. A weekly community band rehearses together. After a rough rehearsal, one frustrated member says to another:
“You’re completely unreliable, you clearly don’t care about this band.”
Your task. Rewrite that line as an observation of behavior plus its effect, with the character verdict removed. To do it, imagine what a camera might actually have recorded that led to the “unreliable” judgment, for example missed rehearsals or late arrivals, and state that, along with the effect it had on the group. One or two sentences is plenty.
Write this in your own words, from your own read of what the behavior behind the judgment might be, rather than asking an AI tool to generate it for you. Turning a judgment into an observation is the skill you are building here. Then check your answer against the model solution found at the end of this chapter.
Let’s Recap!
An observation is what a camera could record, the specific thing said or done, while an evaluation is a judgment about the person.
Evaluations land as attacks and trigger defensiveness, so the person defends their character instead of engaging the problem.
To turn a judgment into an observation, catch the judgment word, find the observable behavior behind it, and state that behavior with its effect.
The camera test settles it: if a camera could not have filmed it, it is a verdict, not an observation.
Describing instead of blaming lets a team raise hard things without burning the trust and safety built in Part 1.
You can now raise a hard truth without turning it into a personal attack. But some of the things that quietly erode trust are not single sentences, they are patterns that run under the surface of a group without anyone naming them. In the next chapter, we learn to spot the everyday barriers that shut voices out, dominating, interrupting, dismissiveness, and in-group habits, and the moves that undo them.
Model Solution
There is no single right wording, since it depends on the behavior you pictured, but a strong rewrite might look like this.
“You’ve missed the last three rehearsals without letting us know, and we’ve had to rework parts on the day because we were a player short.”
Notice what it does and does not do. It drops “unreliable” and “you don’t care,” the two character verdicts, entirely. It names a specific, observable behavior, three missed rehearsals with no notice, that a camera could have recorded, and it states the effect on the group, reworking parts short-handed. It says nothing about what kind of person the member is, which is exactly what keeps them able to hear it. A weaker answer keeps a verdict smuggled inside, “you’ve been really thoughtless about missing rehearsals,” where “thoughtless” is still the camera-invisible judgment doing the damage.
