Recently, I conducted a fun A/B test: when refusing a cigarette offered by someone, what’s the difference between saying “I don’t smoke” and “I’ve quit smoking”? The result, as the title suggests, is that “I’ve quit smoking” clearly makes people more uncomfortable than “I don’t smoke.”
This little experiment wasn’t carefully designed; it was purely accidental. When someone offers me a cigarette and I say I don’t smoke, they usually don’t say much.
But if I say I’ve quit smoking, especially in a group setting, the other person suddenly starts explaining why they need to smoke. If there are more non-smokers present, the person offering tends to justify their “necessity,” such as needing to socialize, being influenced by work environment, or needing to relax due to high life pressure. If there are more smokers present, the person offering tends to show more aggression, questioning the legitimacy of my quitting—for example, asking if I have health or financial issues, criticizing my quitting as “not fitting in,” or saying I’m not qualified to be part of the big shots’ social circle.
If it were just different reactions from different people, I wouldn’t find it that interesting—too many variables to observe. But when I realized the different outcomes of the two responses, I deliberately tested both statements on the same person in different settings, and found similar results. This piqued my interest. Why does the same “fact,” when framed differently to the same person or type of person, elicit different responses?
I submitted my questions and observations to ChatGPT for analysis, and it roughly gave me this explanation:
- When people face situations that threaten their self-justification, they experience “cognitive dissonance” and tend to alleviate it through explanation or defense.
- Individual behavior in a group is influenced by “social norms” and “conformity effects,” especially when there’s a clear majority or minority.
- Framing the same fact differently can significantly change others’ understanding and reactions.
- When one person’s choice is implicitly presented as “superior,” it puts psychological pressure on others’ existing choices.
This explanation has some logic but only covers the “quit smoking” scenario, so I pressed further. GPT explained that when I say I don’t smoke, it expresses a neutral state, indicating I’ve never been in the same category as smokers.
But when I say I’ve quit, it expresses a transition from smoker to non-smoker. This transition implies a comparison and judgment about smoking: “quitting” signifies a negation of smoking behavior. This is the main cause of the other person’s cognitive dissonance, because I’ve negated their self-worth.
When cognitive dissonance arises, the other person needs to self-repair, so they start explaining. When smokers are in the minority, they adopt the role of “the one being scrutinized” and engage in self-justification. When smokers are in the majority, they feel the group norm favors “smoking is normal,” so they’re more likely to show aggression to defend the group’s legitimacy.
GPT further explained that this comparison usually only occurs within the same group. That is, “non-smokers” and “smokers” belong to different groups, so no value comparison arises. But “quitters” and “smokers” may have once been in the same group, and the divergence in values creates the greatest conflict.
When smokers face a quitter, their aggression actually stems from inner vulnerability. When they say “smoking is for socializing” or “quitting means losing opportunities,” they’re defending themselves. The more people there are, the louder this collective defense becomes, because the group provides cover for “the banality of evil,” making them feel this absurd logic is reasonable.
This reminds me of many self-media outlets that often use cognitive dissonance mechanisms to craft content. In stirring up conflict, they use themes like “self-discipline is ruining your life” to challenge mainstream values and gain traffic (haters are still fans); or they dish out chicken soup for the soul, framing “staying up late” as “deeply occupying time” (shoutout to late-night master Zhang Chaoyang). Essentially, they exploit psychological mechanisms, using value conflicts to quickly stir readers’ emotions and convert their negative emotional release into traffic.
Of course, there are also positive uses, such as constructive coaching that helps people face problems, bringing issues to the surface until they must be addressed.
Although my A/B test is crude and can’t support further verification, it already reveals some interesting points: the same psychological mechanism, in different hands and with different goals, can produce different outcomes; many expressions intentionally or unintentionally turn information transmission into an intervention in others’ self-evaluation systems, because stress comes from challenging the other’s self-assessment; expressing the same fact in different ways sometimes changes not the information itself, but the other person’s position within that information.
So most of the time, opinions aren’t that important. What matters more is where one’s self stands in relation to the other’s opinion.