CONCEPT
The theory and limits of owning your outcomes.
Personal responsibility is an allocation rule for attention, not a verdict about causation. It answers the question "who should act on this?" and it gets confused, constantly, with the separate question "whose fault is this?" Those come apart all the time. A situation can be overwhelmingly not your doing and still be entirely yours to handle. Keeping the two questions separate is most of the skill.
Causal responsibility is a question about the world: what inputs produced this outcome. Moral blame is a judgment about an agent: did they act badly given what they knew. Operational ownership is an assignment: who is on the hook for the response. Ordinary language collapses all three into "responsible," and the collapse is where the bad reasoning lives.
The Stoic version, from Epictetus, splits the world into what is up to us and what is not, and puts all effort on the first. Rotter's locus of control is the psychological cousin: the degree to which someone attributes outcomes to their own action versus to circumstance. The correlations run the direction you would guess, and the naive reading of them is still wrong. Ownership belongs to your response function, not to the input distribution. You do not own the market, the illness, the other person's choices, or the year you were born into. You own what you do with the draw.
The first is responsibility inflation: taking everything as yours. It presents as rigor and it is actually a mis-specified model. You have assigned yourself variance you do not control, and because the model is wrong you keep tuning parameters that were never connected to the outcome. The result is high effort, no learning, and eventually the conclusion that effort does not work. It looks like maturity from the outside for about two years.
The second is the attribution asymmetry that Lee Ross named the fundamental attribution error: explaining other people by their character and yourself by your circumstances. He was late, he is unreliable. I was late, traffic was bad. Nearly everyone runs this and nearly no one runs the correction, which is simply to apply the same explanatory standard in both directions and see which of your judgments survive.
The practical test for whether a responsibility claim is real: name the lever. If you cannot state a specific action available to this person that would have changed the outcome, you are not assigning responsibility, you are assigning fault, and fault does not improve anything. Anyone who has run an operations team knows the pattern — someone is held accountable for a number they have no authority to move, and the only thing that changes is that they get better at explaining the number.
The same failure runs inside a single person. Holding yourself accountable for outcomes you have no lever on produces guilt, and guilt is not a control input. It generates no correction, it just consumes the attention that a real correction would have needed.
So the question is not "is this my fault." It is "what is the highest-leverage thing available to me from here." The first question is about the past and has no action attached. The second is the only one that ever pays.