Observations and Evaluations: Why the distinction matters in Nonviolent Communication
- Claudia Sánchez
- 2 days ago
- 8 min read
A sentence can sound factual and still carry a judgment. “You ignored me” feels like an observation, but it usually contains an evaluation about someone’s intention. “You looked at your phone while I was talking” gives everyone more room to respond.
That difference sits at the heart of Marshall B. Rosenberg’s work in Nonviolent Communication: A Language of Life. Rosenberg argues that conflict grows when people mix what they actually saw or heard with labels, blame, assumptions, or predictions. Specific observations do not solve every problem, but they make honest conversation much easier.
This guide summarizes Rosenberg’s main examples, explains the types of evaluations he warns against, and shows how to use the same skill at home and at work.
Why observations matter so much
Rosenberg’s point is not that evaluations are bad. People evaluate constantly. We prefer some behavior over other behavior. We judge whether something feels safe, kind, fair, helpful, or frustrating.
The problem starts when we present an evaluation as if it were a neutral fact.
For example:
Evaluation mixed with observation | Specific observation |
“You never listen.” | “When I was talking about the bill, you looked at your phone and did not respond.” |
“This report is sloppy.” | “The report has three missing citations and two sections that repeat the same point.” |
“The kids are out of control.” | “The kids are running through the hallway and shouting.” |
The difference is practical. A judgment invites defense. A clear observation gives the other person something specific to confirm, deny, explain, or change.
Marshall Rosenberg’s main examples in plain language
In Nonviolent Communication, Rosenberg gives a set of examples that show common ways people confuse observations with evaluations. The examples below are paraphrased and organized by type.
Type 1. Using “to be” as if a judgment were a fact
One of Rosenberg’s first warnings is about sentences like “You are too generous.” The sentence treats the speaker’s judgment as a fact about the other person.
A more specific version takes responsibility for the evaluation.
Mixed statement | Specifict observation |
“You are too generous.” | “When I saw you give away your lunch money, I thought you were giving more than you could afford.” |
The second version still includes an opinion, but it owns it. It does not pretend that “too generous” is an objective measurement.
Family example:
Evaluation | Observation |
“You’re careless with money.” | “This week, you bought three things that were not in our budget.” |
Work example:
Evaluation | Observation |
“You’re unreliable.” | “You arrived after the scheduled start time on Monday and Wednesday.” |
The key move is simple: replace “you are” with what you saw, heard, counted, or can point to.
Type 2. Using verbs that carry hidden judgment
Rosenberg uses examples like “Doug procrastinates” to show how a verb can smuggle in an evaluation. “Procrastinates” does not only describe timing. It implies avoidant or irresponsible behavior.
A more specific statement would describe the pattern.
Mixed statement | Specific observation |
“Doug procrastinates.” | “Doug studied for the exam the night before it was given.” |
That may still concern you, but it is easier to discuss.
Family example:
Evaluation | Observation |
“You keep putting off chores.” | “The trash was still in the kitchen this morning after we agreed it would go out last night.” |
Work example:
Evaluation | Observation |
“The team is dragging its feet.” | “The draft was due Friday, and I have not received it by Monday morning.” |
This type matters because evaluative verbs can sound precise while still being vague. “Stonewalling,” “nagging,” “slacking,” “micromanaging,” and “overreacting” often work the same way.
Type 3. Treating predictions as certain facts
Rosenberg also points out that people often state predictions as though they were facts. A sentence like “She won’t finish the work” may be a belief, but it is not yet an observation.
A more specific version names the basis for the concern.
Mixed statement | Specific observation |
“She won’t get the work done.” | “She said she had not started the work by 4 p.m., and it is due tomorrow.” |
Family example:
Evaluation | Observation |
“You’re going to forget again.” | “The last two times we planned this, you did not put it on your calendar.” |
Work example:
Evaluation | Observation |
“This vendor will miss the deadline.” | “The vendor has missed the last two checkpoints and has not sent the revised schedule.” |
This distinction reduces tension because it leaves room for new information. Maybe the person has a plan. Maybe the circumstances changed. Maybe your concern is still valid. The observation gives you a cleaner place to begin.
Type 4. Confusing worry with objective truth
Rosenberg gives an example about diet and health to show the difference between a concern and a fact. “If you don’t eat balanced meals, your health will suffer” may sound caring, but it presents a fear as certainty.
A specific version says what is actually happening inside the speaker.
Mixed statement | Specific observation |
“If you don’t eat balanced meals, your health will suffer.” | “When I see you skip meals, I feel worried because I care about your health.” |
Family example:
Evaluation | Observation |
“You’re ruining your sleep.” | “You went to bed after midnight four nights this week, and I feel worried when I see you tired in the morning.” |
Work example:
Evaluation | Observation |
“You’re burning yourself out.” | “You worked late three evenings this week and sent messages after 10 p.m.” |
The goal is not to hide concern. The goal is to state concern as concern, not as a verdict.
Type 5. Turning opinion into fact
Rosenberg warns against statements that treat a personal or cultural opinion as if it were an objective description. He gives examples involving groups and property care to show how quickly this can become stereotyping.
A more specific observation limits the statement to what the speaker actually noticed.
Mixed statement | Specific observation |
“Those people don’t take care of their property.” | “I saw snow still on the sidewalk outside that house this morning.” |
Family example:
Evaluation | Observation |
“Teenagers are disrespectful.” | “When I asked for help with dinner, Alex said, ‘Not now,’ and stayed in the bedroom.” |
Work example:
Evaluation | Observation |
“New employees don’t care about quality.” | “The last two new employees submitted files with missing fields.” |
This is one of the most important parts of this article because it affects trust. Broad claims about people often create shame and resistance. Specific observations make repair possible. Also, these type of evaluation can feed hate speech.
Type 6. Using ability words without evidence
Rosenberg uses a sports example, such as calling someone a poor player, to show another trap. “Poor player” sounds like a description, but it is an evaluation of ability.
A specific version names the observable performance.
Mixed statement | Specific observation |
“Pedro is a poor soccer player.” | “Pedro has not scored in the last several games.” |
Family example:
Evaluation | Observation |
“You’re bad at cooking.” | “The rice was burned tonight, and the chicken was still pink in the center.” |
Work example:
Evaluation | Observation |
“She’s bad with customers.” | “During the call, she interrupted the customer twice before the customer finished explaining the issue.” |
This does not mean you can never assess skill. At work, skill assessment is often necessary. The difference is that useful feedback names the behavior and result, not just the label.
Type 7. Using adjectives as if everyone must agree
Rosenberg also gives examples like calling someone ugly. The word sounds descriptive, but it reflects the speaker’s taste.
A more specific version owns the reaction.
Mixed statement | Specific observation |
“Juan is ugly.” | “Juan’s appearance does not appeal to me.” |
Family example:
Evaluation | Observation |
“That outfit is ridiculous.” | “That outfit has bright orange pants, a green shirt, and a purple hat.” |
Work example:
Evaluation | Observation |
“The design is ugly.” | “The design uses five font sizes, three shades of red, and uneven spacing between sections.” |
At work, this distinction can save time. “Ugly” starts an argument about taste. Specific visual details help people revise. Also, these specific observations can then support the expression of your taste. I opens the conversation instead of closing it.
Type 8. Using “always” and “never” as weapons
Rosenberg gives special attention to words like “always,” “never,” “ever,” and “whenever.” These words are not automatically wrong. They become a problem when they exaggerate.
Mixed statement | Specific observation |
“You’re always busy.” | “The last three times I asked to spend time together, you said you did not want to.” |
“You never do what I ask.” | “On Tuesday and Thursday, the laundry was still in the washer after I asked you to move it.” |
Family example:
Evaluation | Observation |
“You never help with the kids.” | “This week, I handled bedtime on Monday, Tuesday, Wednesday, and Thursday.” |
Work example:
Evaluation | Observation |
“You always miss details.” | “In the last report, the client name was misspelled in two places.” |
“Always” and “never” often turn one problem into a character trial. Specific times and actions keep the conversation grounded.
A practical test for separating observation from evaluation
Use this quick test before giving feedback at home or work.
Ask yourself:
Could a camera or audio recording capture this?
Did I name a specific time, place, action, or exact words?
Am I guessing what the person intended?
Am I using a label like lazy, rude, selfish, careless, dramatic, or irresponsible?
Am I using “always” or “never” when I really mean “several times”?
Can the other person recognize the event I am talking about?
If the answer is no, rewrite the sentence.
Try this pattern:
“When I saw or heard _____, I felt _____ because I need or value _____. Would you be willing to _____?”
For example:
“When I saw the dinner dishes still in the sink this morning, I felt frustrated because I need shared responsability. Would you be willing to wash them before you go to bed tonight?”
At work:
“When I saw that the client file was missing the signed form, I felt concerned because I need complete records before submission. Would you be willing to send the form by 2 p.m.?”
The wording may feel stiff at first. The habit matters more than the exact formula.
How to use this in family conversations
Family life creates perfect conditions for mixed observation and evaluation. People are tired. Patterns repeat. Old hurt gets attached to new events.
A few common translations help.
Say less of this | Say more of this |
“You don’t care about me.” | “When you left the room while I was still talking, I felt hurt.” |
“The kids are being impossible.” | “The kids have asked for snacks four times in 20 minutes.” |
“You’re treating me like a child.” | “When you reminded me three times about the appointment, I felt irritated.” |
“You’re selfish.” | “You took the last slice without asking whether anyone else wanted it.” |
This does not make family talks emotionless. It makes emotion easier to hear.
How to use this in work feedback
Work feedback often fails because it jumps too quickly from behavior to identity.
Instead of:
“You are not leadership material.”
Try:
“In yesterday’s shift, you assigned tasks to three people but did not check whether they had the supplies they needed.”
Instead of:
“Your communication is bad.”
Try:
“The message did not include the due date, the file name, or who should approve the draft.”
Instead of:
“You are negative.”
Try:
“In the last planning conversation, you named three risks and did not name any option you would support.”
Clear observations also help managers, coworkers, and employees stay fair. They reduce vague criticism and give people something they can actually change.
Sources used
Marshall B. Rosenberg, Nonviolent Communication: A Language of Life. The examples in this article are based on Rosenberg’s chapter on observing without evaluating, especially his distinctions between observations, labels, predictions, opinions, and exaggerations.
The Center for Nonviolent Communication, writings and teaching materials on the four-part NVC process of observations, feelings, needs, and requests.
J. Krishnamurti, as cited by Rosenberg, is often associated with the idea that observing without evaluating is a high form of intelligence.
The takeaway
Separating observation from evaluation is a small language skill with a large effect. It does not ask you to stop having opinions. It asks you to stop disguising opinions as facts.
The next time a sentence starts with “You always,” “You never,” “You are,” or “They don’t care,” pause. Name what happened. Name what you felt. Name what you need. Then make a clear request.
That one pause can turn blame into a conversation.
Comments