The Selective Laziness of Reasoning

By Neuroskeptic | October 21, 2015 4:19 am

If you could meet yourself, would you always agree with yourself?


You might hope so. But according to a new study, many people will reject their own arguments – if they’re tricked into thinking that other people proposed them.

The paper, published in Cognitive Science, is called The Selective Laziness of Reasoning and it’s from cognitive scientists Emmanuel Trouche and colleagues. By “selective laziness”, Trouche et al. are referring to our tendency to only bother scrutinizing arguments coming from other people who we already disagree with.

To show this, the authors first got MTurk volunteers to solve some logic puzzles (enthymematic syllogisms), and to write down their reasons (arguments) for picking the answer they did. Then, in the second phase of the experiment, the volunteers were shown a series of answers to the same puzzles, along with arguments supporting them. They were told that these were a previous participant’s responses, and were asked to decide whether or not the “other volunteer’s” arguments were valid.

The trick was that one of the displayed answers was in fact one of the participant’s own responses that they had written earlier in the study. So the volunteers were led to believe that their own argument was someone else’s.

It turned out that almost 60% of the time, the volunteers rejected their own argument, and declared that it was wrong. They were especially likely to reject it when they had, in fact, been wrong the first time.

Trouche et al. conclude that

Participants proved critical of their own arguments when they thought that they were someone else’s, rejecting more than half of the arguments. They also proved discriminating: they were more likely to reject their own arguments for invalid answers than their own arguments for valid answers…

These experiments provide a very clear demonstration of the selective laziness of reasoning. When reasoning produces arguments, it mostly produces post-hoc justifications for intuitive answers, and it is not particularly critical of one’s arguments for invalid answers. By contrast, when reasoning evaluates the very same arguments as if they were someone else’s, it proves both critical and discriminating.

Does this mean that human reasoning is broken? No, the authors say – selective laziness might actually be a good thing, as it helps generate conversations:

Opening a discussion with a relatively weak argument is often sensible: It saves the trouble of computing the best way to convince a specific audience, and if the argument proves unconvincing, its flaws can be addressed in the back and forth of argumentation. Indeed, the interlocutor typically provides counter-arguments that help the speaker refine her arguments in appropriate ways…

As a result, the laziness of argument production might not be a flaw but an adaptive feature of reasoning. By contrast, people should properly evaluate other people’s arguments, so as not to accept misleading information – hence the selectivity of reasoning’s laziness.

There’s just one caveat to these results. During a debriefing phase after the test was over, about half of the participants reported that they noticed the “trick” in which their own arguments were passed off as someone else’s.

The other half of the volunteers did not report suspecting anything, and those “non-detectors” were the ones whose data were included in the final analysis. The people who spotted the trick were excluded; these “detectors” tended to be people who had done well on the logic tests.

So a skeptical interpretation of these results is that some of the participants were just not paying much attention, either to the problems or to their own answers.

ResearchBlogging.orgTrouche E, Johansson P, Hall L, & Mercier H (2015). The Selective Laziness of Reasoning. Cognitive Science PMID: 26452437

CATEGORIZED UNDER: papers, select, Top Posts
  • Ibn al-Haytham

    I hope not too much of the taxpayer money was spent on finding out that many people are rather dumb (especially after you select them based on not being very smart).

  • This Old Housewife

    This explains Fox News in a nutshell.

    • Kate K

      I wonder of you understand how unintentionally ironic your comment is. The science indicates both left and right have their own cognition biases.

  • Uncle Al

    enthymematic syllogisms” Thus social paradigms of grinding the productive to aid the deserving, e.g., the deservings’ do-gooders. Cf: Chewbacca defense; Stanley Milgram’s white lab coats; Klimate Kaos; disarming the public to end gun violence (gun-related violence!).

    Rigorously deriving Euclid cannot draw a map of the Earth’s surface that is undistorted, unfolded, or uncut. Euclid dues not describe the Earth’s surface. “Do something!” is meaningless as a class. “Do something useful,” thus Bolyai then Thurston, but not Euclid. When a discipline appears to spout nonsense, it is.

    • qole

      This comment *almost* makes sense. It uses English words and syntax, and it consists of mostly-readable (jargon-heavy) valid sentences. But, even after repeated parsing and re-parsing, I cannot understand what it is actually saying.

    • dejour

      You are right that most people are unfamiliar with that term. But two words are not the same as the meaningless paragraph you produced. Assuming they described the concept in layman’s terms, it is appropriate that they specified exactly the types of logical arguments that were presented.

  • James


  • angesichts

    Such an incredibly irresponsible paper title, and thesis. The selective laziness OF REASONING? As if reasoning itself, qua reasoning, is responsible. A much more accurate title would be, “The Selective Laziness of Particular People Who Don’t Know How to Use their Reason Well.” (like the authors of this paper)

    • Neuroskeptic

      I think it’s implicit that the title is referring to the human activity of reasoning, not ‘Reason’ itself.

      • angesichts

        Yes I think your reading of it is a fair one, but inserting the adjective ‘human’ before every use of ‘reasoning/reason’ in my original comment would not alter the objection.
        To give an analogy:
        If a researcher interviewed 500 people from the year 1700 and found that 99% of them held patriarchal views (e.g., women should not be politicians, should stay in the home and be ruled by men, etc.), it would be completely invalid to write up one’s conclusions in an essay titled, “The Patriarchy of Reason.” (vs., “The Patriarchy of People Who Don’t Use Their Reason Well”)

  • Hugo Mercier

    Thanks for the coverage of our work!

    Regarding the comment at the end about participants not paying much attention, I would like to stress that they were paying some attention. This is strongly suggested by the fact that even the non-detectors (people who didn’t spot the manipulation) were more likely to reject their own bad arguments than their own good arguments. If they had just not been paying attention at all, then they should have been at chance, no?

    • Neuroskeptic

      Many thanks for the comment!

      It’s true that the non-detectors must have been paying some degree of attention, my point is that maybe they were only partially engaged, and that people who were fully engaged (maybe you could give them a financial motive for getting correct answers?) might be more resistant to the effect.

      • Hugo Mercier

        Yes — in this case my guess would be that the manipulation might still work, but we would need to increase the time between the production and the evaluation of the argument. Here it was a few minutes at most, if we increased that to hours or days, more people, and more attentive people, would presumably fail to detect the manipulation. But it’s a more complex experiment to run.

        • Neuroskeptic

          You could speed things up by using a distractor task in between the two phases of the experiment. Something to make people forget about the syllogisms.

          It could be anything, a general knowledge quiz or a math test perhaps.

          • Hugo Mercier

            Thanks, that’s a good idea! In fact, that was our plan if the experiments we ran didn’t work — to increase the time interval so that fewer people would detect the change. But there’s also something striking about the fact that people reject the argument they have just typed…

          • Sam Lee

            Could the Experiment results not be somewhat skewed by the fact that when manipulation of the arguments occurs, the number of responses on screen is always reduced from 5 (4 invalid, 1 valid) to 2 (1 invalid, 1 valid)?
            It seems likely to me that presented with just two options, people are more likely to be able to discern the valid one (and indeed would be much less likely to even try to reason about one of the other different invalid options).

    • Tyson Douglas

      Did you also ask the participants to reevaluate their own answers? The study seems to suggest the participants may have rejected the answers not only because they were from someone else, but also because they were wrong. It would be interesting to see if a person is more likely to validate/agree with a wrong answer if it was their own wrong answer. This is a very interesting domain to study, and there are many angles from which it can be approached, all worth exploring. So to the author of this article, I would say: there’s more than just one caveat to these results.

      • Hugo Mercier

        That’s a good point. As a matter of fact, we have another paper (just submitted) in which we did something similar, using a different paradigm. People stuck with the answer they were attributed instead of adopting what was actually their own answer but was presented as someone else’s. This suggests that it is the arguments that made a difference in the paper discussed above.

        • Tyson Douglas

          Very interesting. So it seems people tend to agree with themselves, even when (perhaps especially when?) they don’t know what they’re talking about.

  • Ste V B

    This is a fascinating discovery.

    But if someone else thinks so I will have to revise that assessment.

  • feloniousgrammar

    Duck season! Rabbit season!

  • Sam

    Hi there, fascinating study! Quick question, could you put the results down to some kind of incubation effect? For example, the participant was unconsciously evaluating the answer they had previously given and so on second presentation, although not aware that it was their argument, the participants were able to see that the reasoning was wrong because they had been unconsciously incubating it since they first gave it. This could explain why it took them some time to see that the argument was wrong, the brain was extrapolating the argument unconsciously in between presentations. They would be more likely to accurately reject their own answers simply because they had already seen them before and so had more time to unconsciously work them out somewhat.

    • Hugo Mercier

      That’s a good point, however, I don’t think this is what’s happening. If there was some incubation, we would expect it to happen both between the 1st phase (intuitive answer) and the 2nd phase (production of argument) than between the 2nd and the 3rd (evaluation of argument). This is not what we observe: all the improvement happens at the 3rd phase. (Also, in the first experiment there is no 1st phase, and all the improvement also happens between production and evaluation);

      • sonia

        Were they logicians? I’m not even sure if logicians would hold the same opinion in just 3 phases. Unless those peeps have a high IQ, to me it just seems you tested their recall ability. Reason why someone chose something in respect to a logic puzzle seems as transitory to me as seeing a black and blue dress from a gold and white dress. I think the pts experience, association prowess, ability to reason etc matters in this case. I think people spend more time on opposing views at the time when confronted with it, because it hasn’t gelled in their brains as much as their own views. I mean should they accommodate or reject these views which contributes to their sense of self? It may also be a self-confidence thing. People not as confident are more accommodating because they’re uncertain of their views. Seems to be lots of psychosocial factors hete. I dunno….

  • Nicolee

    To answer the first question that is asked in this blog, yes I totally think I would agree with myself if I met myself. And after reading this blog I would still say the same thing. This study it did not give me enough reason to think differently. The fact that 60% of the time the volunteers disagreed with some of their previous arguments and declared that it was wrong is not surprising to me, due to the fact that people can in the moment give a quick answer and a half-thought through argument. Half of the volunteers picked up on the “trick” and they still disagreed with the argument even after knowing that the argument made was actually their own. Perhaps it’s because I am just reading this blog for one of my classes but I found the findings of this study to be inconclusive and perhaps other methods should be used to test the hypothesis of “selective laziness of reasoning.”

  • Pingback: I’ve Got Your Missing Links Right Here (24 October 2015) – Phenomena: Not Exactly Rocket Science()

  • John Roth

    I think this is somewhat similar to writing things down and then coming back to them. Something that looked good originally, then later it doesn’t. It doesn’t matter if it comes from myself or another person.

  • Pingback: I’ve Got Your Missing Links Right Here (24 October 2015) | Gaia Gazette()

  • Pingback: 인간의 추론 과정에 있어서의 선택적 나태(Selective Laziness)()

  • Pingback: 人間の推論過程における選択的怠惰 - Eliaszine()

  • sonia

    Could just demonstrate disparate memory retrieval, delayed associations, temporal lags etc. or propensity to self-critique. I’m not understanding how transitive memory states demonstrate a solid logical opinion of anything when your opinions can change in the next second. Good if they left the task with pts for a week or so to strengthen these associations – then test. That way these potential confounds or biases are accounted for somehow.

  • Pingback: La pereza selectiva del razonamiento |

  • Pingback: Why you typically consider individuals who see the world in a different way are flawed()

  • lewin

    I love this finding but with such high suspicion it’s hard to draw solid conclusions from this dataset. I realize this is the era of rapid MTurk studies, but a more solid research design could have split this into two phases a week apart, making it less likely participants would remember their own answers. Between the phases, experimenters could have also cleaned up the answers (punctuation/grammar etc) so that superficial writing cues didn’t remind participants it was their own answer.



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar