I once encountered an idea I have kept returning to: our experience of reality is often less black and white than we would like because human beings negotiate meaning constantly. Two people can hear the same words differently, assign different motives to the same action, carry different histories into the same conversation, and leave with genuinely different experiences of what just occurred.
But experience, interpretation, and occurrence are not the same thing. What I felt is part of what happened to me internally. What I concluded from that feeling is an interpretation. What can responsibly be established about what I actually did is another question entirely, and confusing those categories creates problems that emotion alone cannot solve.
That distinction requires humility because human events are rarely recorded perfectly. Memory is imperfect. People mishear each other, remember different details, misunderstand intention, and sometimes become certain about things that remain genuinely uncertain. Even a recording can establish the words that were spoken without necessarily establishing everything that led to them or what either person intended.
Where facts remain uncertain, certainty should remain limited too. I do not think one person's confidence in a memory gives that person ownership of reality, nor do I think every disagreement about an event proves that somebody is lying. Sometimes two people simply do not have enough reliable information to know exactly what happened.
There are other situations, though, where an occurrence is much less ambiguous. A message was sent. A promise was broken. A call was made. Money was transferred. Someone said something, did something, or behaved in a way that can be established with reasonable confidence.
The difficulty often begins after the event, when accurately acknowledging the behavior creates a new problem. It is no longer only a question of what happened. It becomes a question of what that behavior says about the person who did it.
If I think of myself as kind and I behaved cruelly, something has to move. I can revise my understanding of myself, accept that I am capable of behavior I do not admire, and decide what I need to change. Or I can begin revising the event until the behavior becomes easier to reconcile with the person I already believe myself to be.
The second option is psychologically attractive because it preserves the person I was before the uncomfortable information arrived. I do not have to become someone capable of cruelty if the cruelty can be converted into a reasonable response to provocation. I do not have to examine selfishness if selfishness can be redescribed as self-protection.
This does not usually require deliberate dishonesty. A person can sincerely remember being frightened, genuinely have felt betrayed, accurately describe being overwhelmed, and still gradually build an account in which those true emotional experiences occupy so much space that their own conduct becomes harder to see.
The circumstances get larger. My responsibility gets smaller.
That movement can be subtle because nothing in the explanation has to be false. I may really have been angry. I may truly have felt attacked, abandoned, ignored, humiliated, or convinced that someone did not care about me. The problem is not that those feelings entered the explanation. The problem begins when they start doing the work of exoneration.
“I understand why I did it” can slowly become “Anyone would have done it.” Being overwhelmed can become an argument that no different response was available, and feeling betrayed can become a story in which the other person's conduct assumes responsibility for everything that followed. What began as context begins changing who appears to have acted.
Sometimes that change is justified. Coercion can meaningfully limit choice. Severe impairment, age, neurological conditions, psychiatric symptoms, intoxication, trauma responses, extreme fear, and cognitive limitations can affect perception, control, and responsibility. A serious account of accountability has to leave room for the fact that human agency is not identical in every circumstance.
That qualification matters because accountability should not become a blunt instrument. Circumstances can change how much responsibility belongs to someone, what options they could reasonably perceive, and how their conduct should be judged. What circumstances do not automatically do is make an established event cease to have occurred.
I have used the same sentence for years: “Behavior is a choice. Accountability is accepting responsibility for that choice, regardless of the circumstances surrounding it.” I still believe in the principle, but I also think it becomes stronger when we recognize that choice can be constrained. Accountability has to be intelligent enough to consider those constraints without turning every constraint into an escape from responsibility.
The useful question is not whether circumstances mattered. They almost always did. The useful question is whether the explanation is proportionate to what those circumstances actually affected, or whether the explanation has grown large enough to make responsibility disappear.
That is why the sentence “I understand why I did it, and I still should not have done it” matters so much to me. It allows context and accountability to occupy the same space. The emotion remains real, the explanation remains useful, and the behavior remains ours.
Without the first half, we learn very little. If we refuse to understand why we acted badly, we may repeat the same pattern because we never examined the conditions that produced it. Without the second half, understanding becomes a sophisticated way of making sure nothing inside us has to change.
Accountability should not require permanent condemnation either. An event can be permanent without becoming a permanent identity. A person may have lied without becoming incapable of honesty, behaved cruelly without being condemned to cruelty forever, or caused serious harm and later become someone who would respond very differently.
That distinction matters because shame can make acknowledgment harder. If admitting harmful conduct means accepting that I am now permanently a bad person, I have created a powerful incentive to defend the story instead of examining the behavior. The behavior gets smaller, the provocation gets larger, and my own responsibility becomes increasingly difficult to locate.
There is another possibility. I can acknowledge that I behaved in a way I am not proud of without deciding that the behavior contains the whole truth about who I am. That gives me somewhere to go besides denial or self-condemnation.
Compassion belongs there too. We can understand why someone acted badly without pretending the behavior was acceptable. We can recognize fear, illness, confusion, limited agency, or genuine suffering while still describing accurately what occurred and who was affected by it.
An apology becomes more meaningful under that standard because it does not require the person receiving it to approve the explanation before responsibility can exist. A person should be able to say, “This is what I experienced, this is why I reacted as strongly as I did, and this is what I did.” The first two statements help us understand the third, but they do not need to erase it.
Repair requires the same discipline. Context may matter enormously if two people are trying to understand how a relationship reached a particular point. But repair becomes almost impossible if acknowledgment is always conditional on the injured person first accepting the explanation that makes the behavior feel less serious.
Forgiveness does not require an inaccurate account either. Someone can forgive what happened while remembering it clearly, and someone can believe another person has changed without pretending there was nothing serious to change. Compassion becomes more credible when it can survive an honest description.
The same standard becomes much harder when the behavior under examination is our own. Most of us are capable of describing what another person did with remarkable efficiency. We remember the words, the timing, the broken promise, the insult, the moment they crossed a line, and the specific action that hurt us.
Our own behavior often receives a much more elaborate introduction. Before we describe what we did, we explain what happened earlier, how exhausted we were, what the other person knew would upset us, what they had been doing for weeks, and why our reaction made sense at the time. By the time we finally reach the action, the listener has already been given a detailed argument for acquittal.
When describing what someone else did, we often begin with the action. When describing what we did, we often begin with the explanation. That asymmetry is worth noticing because it reveals how easily context becomes generosity when we are the ones receiving it.
Sometimes the most useful thing we can do is recover the direct verb. I yelled at them, lied about it, sent the message, ignored the request, said something cruel, or broke the promise. Those statements do not tell the entire story, but they identify the part of the story that belongs to my conduct before the explanation begins competing with it.
When describing what someone else did, we often begin with the action. When describing what we did, we often begin with the explanation.
That directness should apply outward too. Other people's behavior deserves context just as ours does, and accountability is not a license to flatten someone else's motives while demanding nuance for our own. The standard has to remain comparable regardless of whose behavior we are examining.
This also leaves room for mixed responsibility. Someone can have treated me badly and I can still have responded badly. My reaction can be understandable without being admirable, and I can have been genuinely hurt without every conclusion I reached while hurt becoming accurate.
None of that requires equal blame. One person may bear far more responsibility for a particular event than another. The point is that another person's wrongdoing does not automatically take ownership of everything I did afterward.
Human beings tend to prefer cleaner stories. We want one person to have caused the collapse, one person to have been reasonable, and one account to explain everything. Mixed responsibility is less satisfying because it leaves us holding truths that do not cancel each other.
That discomfort is another place where revision can begin. If my behavior does not fit comfortably inside the story in which I am entirely right, I can either complicate the story or simplify the behavior. Simplifying the behavior is usually easier.
There is an important distinction here between legitimate reinterpretation and self-protective revision. Our understanding of an event should change when better information appears. If reliable evidence corrects a memory, if missing context reveals that someone lacked information we assumed they possessed, or if a statement we considered malicious turns out to have had a different meaning, changing our interpretation is not weakness.
That kind of revision increases accuracy. The new account is better because new information earned the change.
Self-protective revision works differently. Nothing materially new has been discovered. What changed is that the established account became difficult to reconcile with the person we want to believe we are.
That is the point where explanation can become strangely creative. The behavior that once felt obviously ours begins acquiring passive language. “I said it” becomes “Things were said.” “I sent the message” becomes “The situation got out of control.” Agency migrates away from the speaker because the grammatical distance makes the moral distance easier too.
We are remarkably skilled at doing this without noticing it. Language gives us many ways to make ourselves less visible inside our own conduct, especially when direct description creates shame. That does not mean every softened description is manipulation, but it should make us curious about why certain verbs become harder to say when they belong to us.
The real loss is not simply accuracy. If I revise the story every time my behavior threatens my preferred self-understanding, I remove the information that could have changed me. I preserve the identity by weakening the feedback.
That is why accountability can be less punitive than avoidance. Accurate acknowledgment gives behavior somewhere to go. It can become information about what I need to watch, repair, understand, or change.
Revision aimed at self-protection offers less. It resolves the conflict by making the conduct less threatening rather than making future conduct better.
This is also why I do not think accountability should depend on perfect agreement between the people involved. Two people may disagree about what led to a moment and still recognize that a particular response crossed a line. They may never settle every question about motive, tone, or fairness and still be able to identify conduct that belongs to each of them.
That ability requires giving up the idea that understanding context must always produce exoneration. Sometimes understanding gives us a more compassionate view of a person while leaving responsibility substantially intact. Sometimes it reduces responsibility because agency really was limited. Sometimes it reveals that our original judgment was too harsh. The task is to let the facts determine which movement is justified rather than deciding in advance that the conclusion has to protect us.
That principle matters just as much when we believe we were wronged. Being hurt can make our own account feel morally privileged, but pain does not make every interpretation accurate. Another person can be responsible for hurting us without becoming responsible for every decision we made while hurt.
Likewise, another person's description of our conduct does not become true simply because they feel strongly about it. Where facts are disputed, both people owe the uncertainty respect. Accountability is not accepting every accusation; it is refusing to manipulate the investigation simply because we dislike where an honest investigation might lead.
That is a harder standard than either defensiveness or automatic self-blame. It asks us to remain open without surrendering judgment, to consider context without allowing context to become camouflage, and to distinguish uncertainty from discomfort.
The deepest challenge appears when the facts are clear enough and we still do not like what they reveal. At that point, we have reached the part of accountability that no argument with another person can resolve for us. We have to decide what moves.
Maybe our understanding of ourselves becomes a little less flattering and a lot more accurate. Maybe we discover that kindness was not as consistent as we thought, that our patience had limits we had not admitted, or that under enough pressure we were capable of conduct we would previously have assigned only to other people.
That discovery is uncomfortable, but it is useful. A self-understanding that cannot survive contradictory information is not really self-knowledge. It is a story that requires protection.
The alternative is to move the event instead. We enlarge the provocation, narrow our agency, soften the verbs, foreground the emotion, and keep revising until the conduct once again fits comfortably inside the person we already believed ourselves to be.
That may preserve self-respect for a while. It does not give us much reason to become different.
I think accountability asks for something more demanding and more hopeful. It asks us to believe that an accurate account of our conduct can challenge our understanding of ourselves without destroying us, and that being wrong about who we were in a particular moment does not prevent us from becoming better in the next one.
The choice is not between shame and denial. There is another path in which the event remains accurately described, context receives the weight it deserves, responsibility is calibrated honestly, and the person is still allowed to change.
That is where feelings reach their proper limit. They can tell us what an experience was like from the inside and help us understand how we arrived at a particular response. They can even reveal information that facts alone would miss. What they cannot responsibly do is become the reason an established account of our own conduct must keep changing until the person in the story looks like the person we hoped we were.
When our behavior contradicts our preferred understanding of ourselves, something does have to move. We can keep revising the account until our existing self survives untouched, or we can let an accurate account of what we did alter the way we understand ourselves enough to change what we do next.
The second choice is harder because it offers less immediate protection. It also happens to be the only one that makes accountability capable of producing anything beyond a better explanation for the person we already were.