10/16/2026
The Addiction of Satisfaction
*When feeling finished becomes more important than being right.*
Sometimes, “Now I understand” means, “Thank God. I can stop thinking.”
Those are not the same achievement.
Imagine a project misses its deadline. Someone announces, “People just don’t care anymore.”
Beautiful. We have an explanation, a moral judgment, and absolutely no need to examine the changing requirements, missing resources, conflicting instructions, or possibility that the deadline was invented during a brief but powerful episode of managerial optimism.
Maybe some people genuinely did not care. That possibility belongs in the investigation.
It does not get to replace the investigation.
What interests me is the moment an explanation becomes so satisfying that we stop asking what it leaves out.
Because satisfaction can come from different places. We might have genuinely understood something. We might have found a story that agrees with our existing beliefs. We might simply appreciate having somewhere to put the blame.
Those possibilities need to be separated rather than wrapped together in the phrase, “It just makes sense.” Research on explanatory satisfaction, confidence, and morally charged judgments gives us reasons to take that distinction seriously.
An answer can settle our feelings without settling the question.
And when we stop checking because we enjoy that settlement, satisfaction has started doing a job that belongs to evidence.
-> The feeling of understanding is not a receipt
There is research almost inconveniently well suited to this problem.
In five studies, Emily Liquin and Tania Lombrozo examined what makes explanations satisfying. Satisfaction was closely connected to how much useful learning people believed they had gained. Its relationship with actual learning, measured through tests, was much less clear. Greater satisfaction also predicted curiosity about related questions.
That last finding matters.
Satisfaction is not automatically the enemy of curiosity. It can accompany and encourage further inquiry. The problem is treating the feeling as independent proof that we have learned, or as permission to stop testing an explanation. That is my concern here, not the claim that enjoying an answer makes someone intellectually defective.
“Addiction” is a metaphor in this article, not a clinical diagnosis. I am describing attachment to the payoff of an explanation, not proposing a new disorder.
It would be a little embarrassing to write about oversimplification and then explain every bad argument with “dopamine.” We should at least make it through the introduction before becoming the exhibit.
A more useful concept is *need for cognitive closure:* the motivation to resolve uncertainty and arrive at a definite answer. People differ in how strongly they prefer that resolution over continued ambiguity. This is about wanting an answer settled, which is not identical to wanting it settled correctly.
The distinction is worth keeping plain.
“I have enough evidence to make this decision” is one stopping rule.
“I no longer feel uncertain” is another.
Sometimes they arrive together. We should not assume they always do.
-> Confidence can close the information desk
Consider an experiment that separates *feeling right* from *being right.*
In a 2018 study, Kobe Desender and colleagues created conditions in a perceptual task where people’s accuracy was matched, but their confidence differed. Participants were more likely to seek additional information in the lower-confidence condition, despite being equally accurate. A preregistered replication reproduced the finding.
Their willingness to keep checking tracked something beyond how well they were actually performing.
That does not mean confidence is useless. It means confidence can influence when information gathering stops, and its calibration matters.
Another study, published in 2020, examined how people processed additional evidence after judging the direction of moving dots. High confidence favored processing evidence consistent with the initial decision while sharply reducing the influence of contradictory evidence. This was a controlled perceptual task, not a complete explanation of political arguments, but it demonstrated that confidence can shape how new information is used.
Here is the practical concern I would draw from those findings:
An explanation does not necessarily need to survive much scrutiny to become difficult to dislodge. Feeling confident can affect both whether we keep looking and what happens when additional evidence arrives.
So the question is not merely, “Did I consider the evidence?”
It is also, “What did I require the evidence to do?”
Did supporting information get a friendly wave through security while contradictory information needed three forms of identification and a letter from God?
That is a standard we can examine without pretending to read our own unconscious motives.
Look at which claims you checked. Look at which sources you questioned. Look at whether the standard changed when the conclusion became personally convenient.
The paperwork is more useful than the personality theory.
-> Naming something is not necessarily explaining it
Return to our failed project.
“People are lazy.”
All right. What does that explain?
Which people? Which tasks? Under what conditions? Were they unwilling, unable, poorly informed, or responding to competing demands? Did the same people perform differently when those conditions changed?
Without those details, “lazy” may be little more than a description of the outcome wearing a fake mustache.
They did not do the work because they are the kind of people who do not do the work.
Excellent. We have successfully driven around the block.
A 2017 study by Carly Giffin and colleagues found that explanations of behavior were judged more satisfying when they included a named tendency or condition than when equivalent explanations omitted the name. The researchers also found evidence that a label encouraged people to infer an underlying cause.
That is not an argument against useful categories.
It is a warning about the distance between identifying a pattern and understanding what produces it.
Calling someone “selfish” does not, by itself, establish why they made a particular decision. Calling a workplace “toxic” does not identify which practices need to change. Calling a disagreement “insecurity” does not answer the disagreement.
A label can be a starting point. It becomes a problem when it is treated as the completed investigation.
The same standard applies to explanations that sound compassionate.
“They are struggling” may be true. It still leaves questions about what they can reasonably do, what support would help, and what behavior other people should not have to absorb.
An explanation should not have to choose between condemning someone completely and excusing them completely.
For the project, perhaps unclear instructions contributed to the failure and a team member knowingly ignored a clear responsibility. Those claims can coexist.
Finding a contributing condition does not erase a choice. Finding a bad choice does not erase the conditions.
-> Moral righteousness offers the deluxe package
A morally satisfying explanation can do more than tell us what happened.
It can also assign us a position in the story.
Consider the difference between these statements:
“I believe this proposal will reduce harm.”
“Anyone who questions this proposal does not care about people.”
The second statement attempts to settle a factual disagreement by turning it into a character test.
It also makes the conversation remarkably convenient. We no longer need to establish whether the proposal works. We can discuss whether the person asking about its effects is sufficiently good.
That is a wonderful arrangement for a weak argument.
A 2026 study by Robin Bayes examined how U.S. adults evaluated scientific information and expert recommendations related to policy. Evaluations depended on whether the information supported existing attitudes, and the difference was greatest for attitudes held with strong moral conviction. The clearest pattern was more favorable evaluation of supporting information, rather than consistently harsher rejection of opposing information.
That distinction is important.
The vulnerability was not simply, “People get angry when challenged.”
It also involved being especially welcoming toward information that supported a morally important position. The study did not find convincing evidence that these effects spread into a general rejection of science.
The lesson I take from that is not to care less.
It is to avoid allowing the importance of a goal to lower the standard for claims about how to achieve it.
“We should reduce suffering” expresses a value.
“This particular intervention reduces suffering” makes a claim about outcomes.
The first does not establish the second.
Compassion is not a results section.
Neither is toughness, patriotism, fairness, freedom, or personal responsibility. Those commitments can help *define* what matters to us. They cannot, by themselves, tell us which explanation is accurate or which action will work.
We can be morally serious while remaining uncertain about causes, tradeoffs, or implementation.
In fact, I would argue that caring about the outcome creates a reason to examine those things *more* carefully, not less.
The people affected by a decision have to live with its results. They cannot spend our good intentions at the grocery store.
-> Simple is not the same as simplistic
Here is point at which this very argument could become its own nothing-burger.
It would be easy to conclude that simple explanations are usually wrong and complicated explanations represent deeper thinking.
Ummm… No.
A complicated explanation can be bu****it with a table of contents.
The number of moving parts does not establish whether any of them belong there.
Research also gives us a more flexible picture than “people blindly prefer simplicity.” Across seven preregistered experiments involving 2,820 participants, Claudia Sehl and colleagues found preferences for simpler explanations and simpler methods of producing outcomes. But when participants received information about the reliability or prevalence of the causes involved, those preferences often weakened or reversed.
People were not simply incapable of moving beyond the simpler answer. Relevant information changed their judgments.
The useful distinction is what an explanation leaves out.
-> Simplification removes details we can reasonably set aside. Oversimplification removes details that could change the conclusion. The comfortable explanation can be ugly
Comfort does not always mean optimism.
Consider someone who answers every question about relationships with, “Everyone is selfish.”
The statement sounds bleak. But as an explanation, it can still offer a kind of convenience: no need to distinguish between people, situations, competing needs, or patterns that change over time.
The same issue appears in “Everything is rigged,” “Nobody ever changes,” or “People like that are always the problem.”
I am not claiming that everyone who says these things secretly wants an excuse. That would reproduce the mind-reading problem.
I am asking what happens when these statements become universal answers.
Do they help us make distinctions and test predictions? Or do they allow every possible outcome to confirm what we already believe?
A story can be emotionally painful and still make further investigation feel unnecessary.
Even a harsh self-judgment can do that.
Compare “I am just a terrible person” with “I avoided that conversation, concealed this information, and need to address the consequences.”
The first gives us an identity verdict. The second gives us specific conduct to examine.
Neither kindness nor cruelty toward ourselves guarantees accuracy.
Sometimes the more honest account is less flattering than we hoped. Sometimes it is less damning than we feared.
The target is not to find the story that feels best or worst.
It is to find the account that best survives contact with what happened.
-> Digging deeper can become another hiding place
There is an equally important trap on the other side.
We can keep investigating long after there is enough evidence to act.
Perhaps we keep generating explanations for someone’s repeated behavior because none of the explanations requires us to make an uncomfortable decision.
Perhaps we reject a well-supported conclusion because accepting it would be inconvenient, then congratulate ourselves for remaining “open-minded.”
Those are possibilities to test through our behavior, not motives to assign automatically.
The observable question is whether additional inquiry is changing the decision.
Are we examining genuinely new evidence? Are we comparing plausible alternatives? Or are we reopening the same question whenever the answer requires something of us?
“More thinking” is not automatically better thinking.
An extra hour spent inventing defenses for a preferred conclusion does not become intellectual rigor because we used a notebook.
Likewise, understanding every detail of another person’s history is not a prerequisite for deciding what behavior we will participate in.
You can be uncertain about why someone repeatedly breaks an agreement while being clear that the agreement is repeatedly broken.
You can act on a strong conclusion without pretending it explains everything.
That is the difference between a decision and a declaration of omniscience.
We need decisions. Omniscience seems to be taking longer than the marketing suggested.
-> You are allowed to stop. You are not entitled to certainty.
Nobody has the time or capacity to conduct a doctoral investigation into every claim.
The practical goal should not be endless doubt. It should be a defensible stopping point.
For a consequential decision, I would want to identify the actual claim, examine the strongest relevant evidence I can reasonably access, consider a credible alternative, and name what could change my conclusion.
The depth of that work should match the stakes.
Choosing a sandwich does not require the same investigation as publicly accusing someone of misconduct. Nor does every remote possibility deserve equal attention merely because somebody can imagine it.
A reasonable conclusion can be strong without being absolute.
“The available evidence supports this explanation, with these limitations” is not a failure to take a position.
It is a position that tells us what it rests on.
And sometimes the honest stopping point is simply, “I do not know, and this is not important enough to investigate further.”
That is different from deciding that the first comfortable explanation must therefore be true.
You can close the browser without closing the case.
We should also assess people across time rather than making one hurried answer into a permanent identity. The important question is how someone responds when relevant evidence becomes available.
Do they correct the claim? Narrow it? Acknowledge the exception? Explain why the new information does not actually change the conclusion?
Those responses tell us more about their thinking than whether their first answer arrived quickly.
-> The lighthouse test
The point of this article is not to provide a satisfying explanation of why other people are so easily satisfied.
That would be a very tidy place for the whole argument to collapse.
The useful application is closer to home.
Signal.
State the claim precisely enough that it can be checked. “The team does not care” is broad and interpretive. “These three tasks were not completed after responsibilities were confirmed” gives us a starting point. Then ask what evidence would distinguish unwillingness from confusion, overload, or another plausible explanation.
Mirror.
Notice what your preferred explanation permits you to stop examining. Does it remove your contribution? Make someone else entirely responsible? Spare you a difficult conversation? The benefit does not prove the explanation false. It identifies a place where your conclusion deserves extra checking.
Sovereignty.
Decide separately what you believe and what you will do. You may need to act before every uncertainty is resolved. Choose the action the evidence supports, keep its limits visible, and identify what would justify changing course. Do not confuse a necessary decision with a complete understanding.
Gritty invitation.
Pick one explanation you have used recently that felt especially satisfying. Write down the strongest fact it explains, the most important fact it struggles with, and one realistic piece of evidence that would make you revise it.
*Then, actually look for that evidence.*
Not because your conclusion must be wrong. But because a conclusion you are unwilling to test has begun receiving special treatment.
The project may still have failed partly because someone did not care. The policy you support may still be the strongest option. The person you distrust may have earned that distrust.
You do not have to surrender a supported judgment to prove that you are open-minded.
You do have to let support, rather than satisfaction alone, carry the judgment.
Enjoy the moment when something finally makes sense.
Just do not confuse the relief of finding an answer with the work of establishing it.
The answer that lets you stop thinking has not necessarily earned the right to.
- - - - - -
References:
Bayes, R. (2026). Seeing the same evidence differently: Biased assimilation and moral conviction in public evaluations of scientific expertise. Political Psychology, 47(3), Article e70146. https://doi.org/10.1111/pops.70146
Desender, K., Boldt, A., & Yeung, N. (2018). Subjective confidence predicts information seeking in decision making. Psychological Science, 29(5), 761–778. https://doi.org/10.1177/0956797617744771
Giffin, C., Wilkenfeld, D., & Lombrozo, T. (2017). The explanatory effect of a label: Explanations with named categories are more satisfying. Cognition, 168, 357–369. https://doi.org/10.1016/j.cognition.2017.07.011
Growiec, K., & Szumowska, E. (2026). Need for cognitive closure predicts preference for similar others and reduced diversity in social networks. Scientific Reports, 16, Article 5582. https://doi.org/10.1038/s41598-026-36288-6
Liquin, E. G., & Lombrozo, T. (2022). Motivated to learn: An account of explanatory satisfaction. Cognitive Psychology, 132, Article 101453. https://doi.org/10.1016/j.cogpsych.2021.101453
Rollwage, M., Loosen, A., Hauser, T. U., Moran, R., Dolan, R. J., & Fleming, S. M. (2020). Confidence drives a neural confirmation bias. Nature Communications, 11, Article 2634. https://doi.org/10.1038/s41467-020-16278-6
Sehl, C. G., Denison, S., & Friedman, O. (2024). Doing things efficiently: Testing an account of why simple explanations are satisfying. Cognitive Psychology, 154, Article 101692. https://doi.org/10.1016/j.cogpsych.2024.101692
