Skip to content

Negativity (when applied with rigor) requires more care than positivity.

Tyler Cowen writes:

Avoid criticizing other public intellectuals. In fact, avoid the negative as much as possible. However pressing a social or economic issue may be, there is almost always a positive and constructive way to reframe your potential contribution. This also will force you to keep on thinking harder, because it is easier to take apparently justified negative slaps at the wrongdoers.

This is not my approach, so it might be worth exploring our differences here.

1. Most importantly, there’s division of labor, or the ecosystem. I think it’s good to have some writers who are positive and others who are negative. Each of us has our own style. If Cowen is comfortable being positive most of the time and that works for him, that’s cool.

2. Regarding my own negativity: one thing I’ve found is that often I will start from a positive, constructive perspective but then move to the negative after a series of frustrations.

For example, I thought the newspaper columnist David Brooks had some interesting insights regarding Bobos, Red and Blue America, etc., but I gradually got frustrated at his sloppiness and his refusal to correct his mistakes. Where do you draw the line between making a mistake and flat-out lying? I’m not sure.

This comes up a lot: the beauty-and-sex-ratio research, the Why We Sleep book, Pizzagate, ESP at Cornell, ovulation-and-clothing, etc etc etc. People make mistakes, we discuss these mistakes in a constructive way, and we reach escalating levels of frustration as the promulgators of these errors refuse to consider the possibility they may be wrong.

But that’s all background. Here’s my main point:

3. Negativity (when applied with rigor) requires more care than positivity.

Cowen writes that being positive “will force you to keep on thinking harder, because it is easier to take apparently justified negative slaps at the wrongdoers.”

I disagree.

I mean, sure, negativity without rigor doesn’t require much thought at all. I can type “X sucks” and get a reaction, and this doesn’t take any thought at all.

But once rigor is added to the mix, it’s another story. I’d argue that positivity is in many cases the lazy option, whereas negativity requires one to think harder.

Why? Because if you’re positive you often don’t need to carefully defend what you wrote. Consider some examples:

Beauty-and-sex ratio. When Stephen Dubner wrote about that research in Freakonomics, he was positive, avoiding all criticism. In contrast, when David Weakliem and I wrote about it, we were negative. Being negative took more work! We couldn’t just quote the claims approvingly; we had to explain why we didn’t believe the claims and what the researcher did wrong. This in turn led to useful ideas in statistics.

From a political standpoint, maybe it would’ve been savvy for Weakliem and I to keep a positive tone. But if the goal is thinking harder and ultimately contributing to science, the negativity had a real benefit.

Power pose. This is an example where negative takes (see this from Dana Carney, the lead researcher on the original power pose project, this from Joe Simmons, Leif Nelson, and Uri Simonsohn, and this careful analysis from Carol Nickerson) have been crisp and precise, whereas positive takes (see this from Amy Cuddy, Jack Schultz, Nathan Fosse) have been much sloppier. Again, it seems that, if you say something nice, not much justification is needed, but it you say something negative, you need to back up your claims with careful evidence.

Fair enough: if you want to go against the grain you have to work harder to convince people. My point is that this is the exact opposite of Cowen’s claim that following his advice—“Avoid criticizing other public intellectuals. In fact, avoid the negative as much as possible”—will “force you to keep on thinking harder.”

Why We Sleep? Alexey Guzey put in a few zillion hours writing his accurately-if-impolitely titled post, “Matthew Walker’s ‘Why We Sleep’ Is Riddled with Scientific and Factual Errors.” Guzey criticized a public intellectual! But this was not the easy way forward. Posting a positive review of Why We Sleep?, that would’ve been easy. Or making a few criticisms but surrounding them with politeness, that would’ve been easy too. Writing something so negative—that was hard. Guzey had to back up every point. He was so careful. It reminds me of something that Jordan Anaya, one of the Pizzagate critics, wrote a few years ago: “I know Wansink’s work better than he does, it’s depressing really.” I knew what Anaya meant. The standards for post-publication criticism are often higher than for original work. It’s not enough to show that a claim is unsupported; you have to go into the reasoning and figure out exactly what went wrong (as Guzey showed most spectacularly in this example).

Recall the Javert paradox: critics get attacked, just for being critics. As Simine Vazire wrote, let 2020 be the year in which we value those who ensure that science is self-correcting. If that means criticizing public intellectuals, so be it. Careful criticism requires work, but it’s a public good. More Guzeys, please.

Quick takes

I’ve emphasized above that positivity or negativity has little value when not backed up by solid argument. A positive but unsupported bit of puffery such as “the replication rate in psychology is quite high—indeed, it is statistically indistinguishable from 100%,” or unsupported name-calling such as “methodological terrorist” or “Stasi“—that just pollutes the discourse.

But what if all we have time for is a quick take? What then?

I think a quick positive take or a quick negative take is fine—as long as we make it clear where we’re coming from.

For example, I have no problem at all with an uninformed but positive take such as, “I saw this paper X claiming Y. I didn’t have a chance to evaluate the data and analysis, but the claim seems plausible and potentially important, so I wanted to share it with you.” Or just “Check out this paper X” with a quick link, which implies that I haven’t had a chance to evaluate but I think it could be valuable.

I also have no problem with a quick negative take that is described accurately. For example, “I saw this paper X claiming Y. I’m skeptical because I don’t trust the data [or because I don’t think the effect could be this large, or because any effect would be overwhelmed by noise, or . . .], but here’s the link and you can make your call.” Or just “I’m suspicious of the claims in paper X” with a quick link, which implies that I haven’t had a chance to evaluate but I have some reasons for distrust.

P.S. Thanks to Bob for the above picture of a cat who is stern and negative, but fair.


  1. PeterK says:

    Cowen’s point of view seems more psychological and social than scientific. From a scientific point of view, there is no negative and no positive, but people may be more comfortable with positive views.

    It also seems to me to be a rather american comment. The ideas of “positivity” or “negativity”, that it is somehow good to be baselessly optimistic and irremediably nice, are not as present in other parts of the world.

    • Martha (Smith) says:

      I agree with PeterK’s first paragraph. Not so sure about the second paragraph (I don’t have adequate information to be able to agree with it nor to be able to disagree with it. I would be interested in hearing what information others can supply, either pro or con.)

    • Hungarians are generally more negative than Americans–in terms of willingness to critique something thoroughly or to express skepticism over its prospects. Sometimes this goes to extremes, especially when the critic assumes too much authority, but overall I find it refreshing. You can point to a problem in something, and that is treated as normal rather than hostile.

      • Martha (Smith) says:

        “… but overall I find it refreshing. You can point to a problem in something, and that is treated as normal rather than hostile.”

        Reminds me of teaching future teachers, and needing to be very careful to make a distinction between “critiquing something” (with the aim of helping to improve it) and “criticizing someone” (with the interpretation of being rude).

  2. I agree with Andrew’s take here (as could be predicted; I mean, I am an avid reader of this blog). What it says to me is that positive-vs.-negative is orthogonal to the real axis that matters, namely careful-vs.-not. If you’re careful-and-positive, great; likewise if you’re careful-and-negative.

    • The only thing I’d add is that it’s nice when someone does admit that they made a mistake. I’ve had to do that in order to redeem myself and become a more grounded person.

    • Martha (Smith) says:

      I Agee with Steve’s comment.

      I think Sameera’s response is too weak — It’s not just nice when someone does admit that they mad a mistake — from a scientific or other intellectual perspective, admitting one’s mistakes is essential.

      • Martha,

        I wasn’t meaning ‘nice’ to mean simply ‘polite’. Getting someone to admit a mistake is a challenge. You have to crank it out of some people. In some cases, ‘cognitive dissonance’ and ‘groupthink’ overwhelm the potential to admit mistakes.

        I also posted the following:

        Being provocative and angry at someone who is trying to pull a snow job. Taking them to task with a composed manner works in some cases; in other cases, it is worth giving such a person a run for his money.’

        Basically most of us appeal to authority in most contexts.

  3. Andrew,

    You analyzed your own response patterns very well. I came to that same hypothesis a year ago. Tyler Cowen naturally wants to shore up the expert’s gatekeeper role, from my listening to his book-talk at Politica & Prose. That’s fine. We don’t have well-developed avenues for making expertise accountable. I gleaned that from attending the Congress on Peer Review online. I think it was sponsored in part by Stanford University. Anyway, peer review has its problems as we have discovered in recent years.

    I think the tone and attitudes of experts toward rival experts are material for satire. I actually don’t care if I make a fool of myself. So I’m empathetic toward others who do. LOL Tyler Cowen is a fine expert. But I find myself disagreeing with some of his views.

    Then again, I think outsider perspectives while integral are often presented in simplistic ways which undermine them. So the trick, imo, is to seize the moment in a entertaining and substantive way.

  4. paul alper says:

    Just when I was nodding in agreement,this hove into view:

    “From a political standpoint, maybe it would’ve been savvy for Weakliem and I to keep a positive tone. But if the goal is thinking harder and ultimately contributing to science, the negativity had a real benefit.”

    Objectively, the objective case is nominally under attack, especially on NPR programs, but to see it in writing is discouraging.

  5. Yes! As a habitual writer of comments and critiques, I’ve found that you have to be extremely careful in preparing them. Editors are not eager to publish critiques of things they’ve chosen to publish. Fraudsters are quick to resort to lawyers to counterattack. Knowing that you could lose every cent you’ve saved if a jury is impressed by one small tangential slip-up focuses the mind.

  6. Christian Hennig says:

    I’d say there’s a value in being positive and polite while being negative, because occasionally that makes it easier to accept the criticism and increases the probability of constructive debate and admitting errors.

    • Martha (Smith) says:

      Agreed (with the interpretation that here “positive” means being as positive as possible without being dishonest, and especially being positive in tone, and “negative” means “criticizing”).

  7. MM says:

    The required rigour also depends on the follow-up. It is relatively easy to spot problems, not so easy to suggest solutions to them — perhaps that’s another difference in perspective from Cowen. Positivity has a natural rigour cap, whereas negativity may be easier at first, but requires more rigour as you try to identify solutions.

    • Brian says:

      Negative feedback has a positive value to the receiver, in the absence of suggested solutions.

      • Curious says:

        Potentially. This is situationally specific. If the task is well defined and movement outside that well defined path cannot improve the process, negative feedback has a positive value. If, on the other hand, the feedback is a misdiagnosis of the problem or mischaracterization of the issue, the value will, at best, be neutral and potentially negative.

    • jim says:

      “The required rigour also depends on the follow-up. It is relatively easy to spot problems, not so easy to suggest solutions to them “

      I strongly disagree. Most of the time the solutions are easy to know.

      The problem is that they require a lot more work – or make the objective unattainable – so people don’t want them.

      • Martha (Smith) says:

        I disagree with Jim, although I don’t entirely agree with MM either. Sometimes it is one (“not so easy to suggest solutions to them”), sometimes the other (“Most of the time the solutions are easy to know.) Perhaps in different subjects/situations one or the other is predominant.

  8. Oliver C Schultheiss says:

    Hi Andrew,

    you write “I can type “X sucks” and get a reaction, and this doesn’t take any thought at all”. If that’s a kind of flat-out negativity (which you rightly reject), also consider the possibility of toxic positivity, a term that Mary Trump is using in her book on her uncle. Just as you can say “X sucks”, you could say “X is great!” and with equally disastrous consequences in the long run, particularly with regard to the role thoughtful critique of scientific work has in advancing science.

    Personally, I think you’re doing a really good job being fair in your critiques of others’ work. And a big part of that is that your critiques are almost always educational for readers. That’s not something you would get from flat-out negativity or toxic positivity.

  9. almost_asleep says:

    Hi Andrew, while I agree with you in essence, I don’t think Cowen is wrong in his own terms. His goal seems to have a long shelf time and obviously it’s helpful to be uncontroversial. Sometimes one has to be Andrew Gelman to get away with criticizing others. You mentioned the Javert paradoxon. Most people see criticism as a form of virtue signaling in front of one’s peers. They have heard about the criticism in general terms for 20 years or so and are “just not convinced”. They expect one to not only criticize but to do better (which might be impossible). Otherwise they see it as shaming their work.

  10. Anoneuoid says:

    Definitely easier to be negative. Same is it is easier to destroy something than create it.

    When I first figured out what NHST was I realized the thing to do was proper science. Ie, come up with a set of reasonable (yet simplified) assumptions from which I could derive computational models to explain the data I could see. Next get similar datasources from elsewhere as a kind of pseudoreplication to make sure it could generalize at least a bit. Then of course make predictions about future data in the hope someone would check it one day.

    Gave the presentation and got asked “but is the difference significant?” That is when I finally decided not to continue with medical research. Doing things correctly just doesn’t matter when doing the wrong thing is so institutionalized.

    • Martha (Smith) says:

      Anoneuoid said,
      “That is when I finally decided not to continue with medical research. Doing things correctly just doesn’t matter when doing the wrong thing is so institutionalized.”

      Sad, but all too often true.

  11. Martin says:

    Maybe I am missing something, but Cowen seems to be entirely oblivious to what he is permanently doing himself: He has an anal-retentive obsession with ranking everything, up and down – it is even a standard question to his interview partners in his Conversations. I mean, just look at today’s Bloomberg column: He cannot even laud UK’s covid vaccine effort without dumping on countries that fared relatively well with containment (hitherto). I get it, this his slightly contrarian-provocative way of “putting things into perspective”, but it is hard to find another “public intellectual” who is consistently negative without even being honest about it to himself (?) or their readers:

    “It is fine and even correct to lecture the British (and the Americans) for their poorly conceived messaging and public health measures. But it is interesting how few people lecture the Australians or the South Koreans for not having a better biomedical research establishment.”

    It is like ‘Look, dear readers, because the anonymous public judges so *inconsistently* I need to go negative – or to be precise, marvel at their relative failure to go negative.’ Is he a child?

    Column is here:

    • Tyler Cowen appeals to people who think as he does; they want to shore up their gatekeeper role in academia. I think Richard Posner is perhaps one of the few that can give Tyler Cowen a worthy debate. Celebrity academics should be held accountable for their claims. We learn on social media that subsets of celebrity academics are human too. Again when one has knowledge of departmental politics, as some graduate students do, one learns of the insecurities of academics quite well. They play into the construction of knowledge as well.

  12. jonathan says:

    So, someone offers an affirmation of a hypothesis which you sensibly treat as having a null, which you investigate to test the affirmation. Sounds pretty reasonable. You could, of course, sloppily investigate, and I think that is where Cowen’s point intersects: most people sloppily investigate, either at the mental modeling level or in details, so you might as well be positive. That is, if you’re constructing a set which includes these ideas, they would tend to find a union the more lack of care is invoked. That transforms this into a question of etiquette; might as well be nice or even dont say anything if you dont have anything nice to say, with the gloss that this includes not caring enough to say something not nice with care.

    • Mendel says:

      I’m trying to remember the general etiquette advice that something you say has to be either “nice” (or useful or beneficial) or true (inclusive or), meaning higher standards apply to negative messages. Which is basically Andrew’s point, and also libel law.

      • jonathan says:

        It isnt libel law unless it crosses a line defined by things like asserting a woman is unchaste, that someone has a communicable disease, etc. Fairly narrow category in the US, larger elsewhere.

        I’m interested in the way academic writing mimics courtesy forms. One interesting part is how courtesy forms mask a lack of underlying rigor by presentation of deferential forms, where deferential is a bow or curtsy (show some leg, me lad). Like in the painting L’Eminence Grise by Gérôme.

  13. jim says:

    Tyler also said:

    “10. …also have some ornery friends determined to make (intellectual) life difficult for you…”

  14. More Anonymous says:

    The trick with being unusually negative is that it also requires one to be unusually self critical. Otherwise, one ends up being not only especially negative, but also especially hypocritical.

    Few critics seem capable of being equally self critical. Instead, they end up criticizing others for issues that, when seen from a distance, appear similar to the mistakes they themselves make.

    Possibly, this is because in many fields (including statistics), bad scientific practices are so entrenched that it is difficult if not impossible to succeed in one’s career without partaking in some of them.

    Also, even if you avoid most bad practices, many colleagues and friends will partake in them. So if you are a harsh critic, you will either have no one who wants to work with you, or you will excuse errors as learning experiences if they occur among friends while deriding similar errors when made by others.

    The result is that the field of public intellectuals seems to be populated by critics who are unaware of their hypocrisy, positive individuals with blind spots to obvious entrenched errors in their fields, and few others.

    To me, this feels like a fairly good state for science to be in! Science has so many wondrous discoveries each year and so many entrenched bad practices, too. In this context, I can’t think of a better composition for public individuals than a mixture of positive people with blind spots and critics with hypocrisies. One is useful to motivate the discoveries and the other to change the bad practices. Problems get better when the balance between the two groups is suitable, and worse when that balance is unsuitable.

    A lot of armchair speculation in this comment, so please take it with salt.

    • Mendel says:

      “Instead, they end up criticizing others for issues that, when seen from a distance, appear similar to the mistakes they themselves make.” — that works well as a general rule for me in Internet debates: that people have been (or are) doing what they’re criticising others for. I suspect this is because the fact that they get criticised for it by others demonstrates to them that a) this type of criticism is socially acceptable, b) there is social status to be gained by levelling it, c) “it’s only fair to hold others to this standard if I have to abide by it”.

      It probably works differently from a teacher perspective, deriving your criticism from seeing lots of other people’s mistakes, but then you’d be more responsibly offering advice for growth along with the criticism.

    • Martha (Smith) says:

      Interesting armchair speculation from MA.

  15. David Marcus says:

    The important criterion is not whether it is positive or negative, but whether it is correct. Categorizing something as positive or negative is usually done by a superficial examination of which words are being used.

  16. Morris39 says:

    I’m an occasional reader not a poster and originally from a field far from economics (or any humanities). The following snippet “Taking them to task with a composed manner works in some cases; in other cases, it is worth giving such a person a run for his money” caught my attention. My filter is that if you express anger/outrage/whatnot you are not in control of your emotions (which don’t interest me) I assume you probably cannot control your case. In my former world being wrong became almost immediately apparent so attacking people is obviously pointless.

    • Curious says:

      If you think it is a “tell” then you can use it to your advantage. However, you have to understand what generates the frustration. If someone consistently uses what they know frustrates another in effort to generate the observed frustration, you begin heading down the road of lawyers behaving badly.

      Bad science and poorly reasoned inference seem to aggravate many on this blog. The only thing that will likely reduce that frustration is more consistently good science and inference published in journals.

  17. Winston Lin says:

    Andy, thanks for this enjoyable post! I think you made some really good points, but I’m not sure Cowen meant what you thought he meant. Here’s a great critique from Agnes Callard:

  18. R. S. says:

    Regarding your third point (and the much of the substance of what you say): “Negativity (when applied with rigor) requires more care than positivity.”

    This is a skewed premise. Surely you should be comparing “negativity (when applied with rigor)” to “positivity (when applied with rigor)”.

    It seems to me that you’re arguing for rigor, with an unstated assumption that only “negativity” can be rigorous. That may or may not be true, but you haven’t demonstrated that here.

  19. Jane Barabe says:

    REaders of serious fiction are pretty much in agreement that book reviews have become appallingly positive. They find ways to like books that are careless and terrible in all sorts of ways. Speculation is that reviewers fear that if they say anything negative at all, even in the midst of a generally positive review, they will be made to suffer.

  20. R.R. says:

    I read Cowen’s quip as saying something slightly different.

    Most notably, I’m not certain that an endorsement of positivity should be interpreted as an endorsement of praise.

    Consider a positive claim to be one where X is true, and a negative claim to be one where X is false. Given these operating definitions, I’m not certain there is any correspondence between negativity and “meanness,” or positivity and praise. In this context, positive claims (or counterclaims, in the vocabulary of debate) certainly require more difficult thinking than plain denial.

    It is easier to make the negative claim, “X is false” than it is to say, “With regard to Claim X, and in light of new evidence, actually, it is Claim Y that is true.”

    Positive claims are more difficult because they require more evidence! But it doesn’t mean we have to forgo criticism of earlier claims in favor of blind endorsement. It only means that criticism, alone, is less constructive than new ideas. (It also doesn’t mean we are allowed to be unnecessarily mean.)

    One benefit from the rule of positive claims, although Cowen never implies it, is that it prevents miscommunication with an underinformed public. Overzealous criticism of well-founded expertise can erode trust in science or intellectualism, broadly, as opposed to creating skepticism about a specific claim. (“Why should I listen to scientists? Even the experts don’t know which way is up!”) When new claims must be derived from new discoveries, we remind the public that science inherently reflects the state of nature, not a state of opinion.

Leave a Reply