I became interested in blaming, because, being part of a ‘normal’ family, I have been blamed for my fair share of things, by almost everybody at some point. But rather than washing dirty linen in public, and inadvertently playing the blame game myself, let’s talk about Bob.

A man I once coached — call him Bob — sat across from me and explained, calmly, why his marriage had failed. His wife had grown cold. Her coldness, he said, traced back to her mother, who had never once told her she was proud of her. Her mother's coldness, in turn, traced back to a father who drank, and the drinking traced back to a war that had chewed up a generation of men and spat out husbands who could not say a soft word to their daughters without flinching. Bob told this story with the satisfaction of a man who has found the true, hidden reason he now slept in the spare room. He had gone back four generations to get there.

He was not wrong, exactly. Every link in that chain was probably real. What he had not noticed — what almost nobody notices while they are doing it — is that he had also stopped. He had walked the chain of causes back through a war and a bottle and a silence passed mother to daughter like a family ring, and then, at some entirely arbitrary point that happened to exonerate him completely, he had simply got off the train.

He had, in short, found the beam holding up the whole structure of his innocence, and stopped digging the moment he found it. Any further and he would have hit himself. Once you see it you cannot stop seeing it: causes really do recede without a natural stopping point. Blame does not. Blame is a stopping rule dressed up as an explanation.

The chain doesn't end. You do.

Follow any effect backwards and the causes fan out and recede, generation by generation, decision by decision, all the way to conditions nobody chose and nobody could have chosen — genetics, geography, the particular temperature of a particular war - the proverbial butterfly that flapped its wings. There is no natural terminus. The chain of causation is, as far as anyone has ever demonstrated, infinite in the relevant sense: every cause has a cause, all the way down, and physics has never once obliged us by putting up a wall.

Human attention, mercifully or not, works nothing like this. Nobody actually walks the chain to its origin, because its origin is nowhere and everywhere and would take longer to locate than the lifespan of the person doing the walking. A 2026 study by Kardos and Lickel, tracing how people assign responsibility in spreading action sequences, found that most of us trace origin back only to roughly the second degree of separation before we treat that node as the beginning of the story. Two steps. Not the true cause. Not even a particularly deep cause. Just deep enough to feel like we've done the work - and we safely say ‘it’s not me’.

The same research found that the harm caused by an initiator multiplies as it spreads — one bad decision cascading into many — we load extra blame onto that first mover, well beyond their actual share of the causal pie. We punish origins for the sins of their descendants. Which tells you something important: the chain is not infinite in the mind. It is truncated, and truncated at a depth that happens to be socially and psychologically useful, not at a depth that happens to be true.

Bob's four-generation excavation, in other words, was actually unusual. Most people don't even make it that far back before finding somewhere comfortable to stand.

What "stopping" is actually for

There is a question underneath the question: if causes recede forever and blame does not, what is doing the stopping? Not truth, because truth has no opinion about where you park the story. Something else is choosing the exit ramp, and it is choosing on the basis of what the story needs to do for the person telling it.

There are a handful of reliable exit ramps, and once you know them you will start seeing them everywhere — in office post-mortems, in your own arguments, in the news cycle's daily search for a single named villain. Let’s explore a few:

A person, not a system. This is the most well-trodden ramp of all. The fundamental attribution error: our stubborn tendency to explain other people's failures by their character and our own by our circumstances, that combines with what safety researchers sometimes call blametropism. It is a gravitational pull, in any investigation, toward the individual standing closest to the failure when it occurred. Charles Perrow, writing about industrial accidents, noted that once an inquiry reaches "the operator failed," it simply stops there. Not because the operator was the deepest cause. Because the operator was the findable one. Every system that produced that operator, including the training, the roster, the fatigue, the twelve prior near-misses nobody acted on, recedes into the fog the moment a human face is available to absorb the story.

I have sat in enough post-mortems to know this ramp by the sound of the chairs. Something goes wrong; a shipment misses a deadline, a client walks, a spreadsheet costs someone six figures, and the meeting has a shape to it that never varies. Twenty minutes are spent establishing who touched the file last. Perhaps five are spent asking why the process allowed one tired, undertrained, under-slept person to be the last line of defence against a six-figure mistake in the first place. The five minutes always feel like a digression, but are, in fact, the only part of the meeting that will prevent a repeat.

Control plus intention. Aristotle got here roughly 2,400 years before the psychology journals did: blame requires that the person could have known and could have done otherwise. Mark Alicke's culpable-control model is the modern, better-tested version of the same idea. We reserve real blame for the node in the chain where we can see a mind that had a live alternative and chose against it. This is why distal causes, the childhood, the economy, the genes, feel less blameworthy than proximate ones even when they are demonstrably more causal. They fail the control test. You cannot glare menacingly at a recession.

Self-absolution. Self-serving bias pushes the chain outward until the story arrives, conveniently, at a point that is not you. Actor–observer asymmetry does the identical trick with a different label: my failure was the situation, yours was your character. But it would be too neat to say the chain always moves away from the self. Sometimes people stop it deliberately close, because they need a villain within arm's reach. Someone punishable, someone real, someone who isn't "the culture" or "the system" or some other abstraction you cannot actually confront at a dinner table. Distance, in other words, gets used both ways: to dump responsibility, and to keep it exactly where it's wanted.

Just-world closure. Melvin Lerner's decades of research on the "just world" hypothesis showed something genuinely strange: people would rather believe a victim somehow caused their own suffering than accept that the world hands out pain at random. This is why the chain sometimes runs the other direction entirely (lengthening rather than truncating) reaching all the way back to karma, or a past life, or a divine plan, rather than stopping at the unbearable answer: chance. If chance is intolerable, people will walk the chain backwards indefinitely to avoid landing on it.

Actionability. This is the quiet, practical ramp, and it is probably the one that governs more of daily life than any of the others. Actual investigators tend to stop at the cause they can actually do something about. It is why organisations blame the last person who touched the broken system rather than the incentive structure that made the breakage inevitable. It is why therapy culture, at its laziest, blames parents: parents are a fixed, namable, blameable category, even when the truer story runs through economics, lineage, culture, epidemiology and a dozen forces no one in the room can move. It is why political culture, at its laziest, blames "the system": an abstraction so large it can absorb any amount of blame without anyone ever having to do anything differently. Different institutions, notice, have different legally and culturally sanctioned stopping depths — a courtroom stops somewhere quite different from a Sunday sermon, which stops somewhere quite different from a boardroom post-mortem.

Whose personality decides where the ramp is

There is no clean diagnostic category for "the kind of person who stops the chain too early". It would have been neat, and I went looking, but psychology has not handed us a tidy blame-typology with a DSM number attached to it. But there are stable styles, and Bernard Weiner's old attribution framework still does most of the organising work: whether we locate a cause as internal or external, stable or unstable, controllable or not.

Depressive explanatory style takes failure and drags it inward, then makes it stable and global. Not "I misjudged this one negotiation" but "I am the kind of person who always ruins things." 

Externalising styles (including some narcissistic and antisocial patterns) do the mechanical opposite, dumping causation outward with impressive speed and confidence. I find it genuinely interesting that fragile self-esteem and grandiose self-esteem can produce the identical outward-blaming behaviour for exactly opposite reasons. The fragile person cannot bear one more confirmation that they are worthless, so the chain must end somewhere else. The grandiose person cannot bear a single crack appearing in the statue, so the chain must also end somewhere else. Same exit ramp, opposite engines.

Culture does something related but distinct. Collectivist cultures, on average, travel further into situation and relationship before calling the chain complete; individualist cultures stop earlier, at the person. (I'd even argue the entire Western legal tradition's obsession with individual mens rea  - the guilty mind - is this cultural stopping-rule, formalised and given a Latin name.) Put people from both traditions in the same room discussing the same failure and the cultural stopping-rule will look, to each side, exactly like a personality flaw in the other.

A typology of the nodes people are willing to blame

If you strip this down, the real question is never "what caused it." It's "what kind of node are we, in this instance, willing to treat as an ending." There is a short list of candidates, and every argument you have ever had about who's really at fault has actually been an argument about which of these should count. 

Which one is your personal favourite?

The proximal agent — whoever last touched it. 

Character — who they fundamentally are. 

Intention — what they meant to do. 

Negligence — what they should have foreseen and didn't. 

System — the incentives and design that made the outcome likely regardless of who was standing there. 

History — the forces that made the agent who they are. 

Chance — the answer almost nobody wants, because it offers no one to punish and nothing to fix. 

Fate, or God — a metaphysical full stop, satisfying precisely because it is unarguable. 

Self - rarest of all, the blamer's own contribution, which culture almost never rewards anyone for reaching.

Watch how differently these nodes are treated by the same institution in the same week. 

A courtroom will happily stop at intention — did the accused mean to do it, did they know what they were doing — and will treat "but the system made it likely" as, at best, a mitigating footnote rather than a defence. 

A safety regulator, investigating the identical category of harm, will do the precise opposite: intention barely registers, and the entire inquiry is built to reach "system," because a regulator's job is to change incentives, not to punish souls. 

A priest, hearing the same story in confession, may reach past both of those and stop at history, or at nothing external at all — at the penitent's own will, freely exercised, full stop. 

None of these three institutions is wrong about causation. They are simply built to extract different jobs from the same chain, and each one has quietly decided, as a matter of professional design, which node counts as an ending.

And that explains something that otherwise looks like hypocrisy: why the same person can, in good conscience, blame the system when they're the one under investigation and blame the individual when someone has wronged them. It isn't inconsistency. It's two different jobs being asked of two different chains, and each chain obediently produces the stopping point its job requires.

Catching the click

Here is a small, practical test, and I use it on myself more often than I'd like to admit. The next time you feel a chain of causation resolve — that little internal click of "ah, there it is, that's what caused this" — stop and ask what would happen if you kept walking one link further back. 

If the honest answer is "the story would fall apart, or implicate me, or lead somewhere I don't want to go," you have not found a cause. You have found a fence, and you built it yourself, and you built it exactly where it was convenient. 

If the honest answer is "it wouldn't change anything useful," you have probably found something closer to a genuine stopping point — the level of the system, say, where an actual lever exists to pull. The test doesn't tell you the truth about causation, because nothing short of omniscience tells you that. It tells you whether the story is serving you or serving the situation, and those are not always, or even usually, the same thing.

I'd even argue that this is the entire difference between a useful post-mortem and a useless one, at any scale — a marriage, a business, a war. The useless kind stops the moment a satisfying villain appears. The useful kind keeps walking past the villain, not indefinitely, but at least one link further than feels comfortable, because the link past comfortable is usually where the actual lever is hiding.

None of this is easy to do in the moment, and mea culpa. I know the mechanism I have just described, and I still catch myself stopping the chain exactly one link short of my own contribution. Knowing the trick does not disable the trick. It just means that, on your better days, you notice yourself pulling it.

What your question is actually asking

So: is the daisy-chain intuition right as metaphysics and wrong as psychology? Yes. Precisely that. Causes genuinely do recede without a natural stop — that part is not folk wisdom, it is close to a description of how causation works. But nobody, anywhere, ever actually reasons their way to the true origin of anything, because the true origin of anything is the Big Bang, or close enough, and that is a useless place to end a sentence about who forgot to pick up the milk.

What people are doing instead is selecting the node that performs the job the moment requires. If the job is innocence, the chain gets walked out far enough that the walker disappears from it entirely, vanishing into "well, that's just how I was raised" or "that's the economy for you." If the job is justice, the chain stops hard at a face,  a single, findable, punishable human being. Because justice as most people practise it is not really interested in causation at four removes, it wants a name. And if the job is learning — actual, useful, non-performative learning — the chain should not be allowed to stop at a face at all, because a face is almost never where the fixable problem lives.

We can’t and need not stop assigning blame, because blaming is useful: it locates responsibility, it deters repeat offences, it lets a wronged person feel the wrong was seen. What I am telling you is to notice, the next time you feel a chain of causation click satisfyingly into a stopping point, that the click is not coming from the evidence. It is coming from you. Ask what the story needed to do, right there, at that exact link, and you will usually find your answer before you find the truth.

That is a harder discipline than it sounds like on paper. It means treating your own sense of resolution as a piece of evidence about your psychology rather than a piece of evidence about the world. Most of us were never taught to distrust our own relief. We were taught the opposite: that when a story finally makes sense, the sense-making is proof of its truth. 

It rarely is. And I mean really rarely. It is usually just proof that the story has, at long last, stopped needing anything more from you.

You cannot forgive a cause. You can only forgive a person you've found to blame. Nobody has ever forgiven "the war," or "the economy," or "the system". Those words don't have a face to look at you while you say it, and forgiveness needs a face. Which means the stopping point isn't only where blame lands. It's also the only place forgiveness can ever happen. Stop the chain at the system, and there's no one there to let off the hook. Stop it at your mother, or your ex, or yourself, and suddenly there is. I suspect that's the real, unadmitted reason the chain gets walked exactly as far as it does and not one step further in a great many families; not because that's where the truth ran out, but because that's where someone forgivable was standing. Which raises the question - what will you do about that?

Bob, for what it's worth, never did go looking for the fifth generation back. He also, tellingly, never went looking for the zeroth — the version of the story that started with what he, personally, had done that same week to make a cold house colder. The chain went back four wars and stopped exactly one link short of a mirror. That is not a coincidence. That is the whole mechanism, working exactly as designed.