When to Parrot, Pander, or Think for Yourself

Humans are built to argue and persuade. We tend to win when we endorse arguments that others accept, and win even more when we can generate new arguments that others will accept. This is both because people notice who originated the arguments that they accept, and because this ability helps us to move others toward opinions that favor our policies and people.

All of this is of course relative to some community who evaluates our arguments. Sometimes the larger world defers to a community of experts, and then it is that community who you must persuade. In other cases, people insist on deciding for themselves, and then you have to persuade them directly.

Consider three prototypical discussions:

  1. Peers in a car, talking on the path to drive to reach an event where they are late.
  2. Ordinary people, talking on if and how black holes leak information.
  3. Parents, talking on how Santa Claus plans to delivers presents Christmas eve.

In case #1, it can be reasonable for peers to think sincerely, in the sense of looking for arguments to persuade themselves, and then offering those same arguments to each other. It can be reasonable here to speak clearly and directly, to find and point out flaws in others’ arguments, and to believe that the net result is to find better approximations to truth.

In case #2, most people are wise to mostly parrot what they hear experts say on the topic. The more they try to make up their own arguments, or even to adapt arguments they’ve heard to particular contexts, the more they risk looking stupid. Especially if experts respond. On such topics, it can pay to be abstract and somewhat unclear, so that one can never be clearly shown to be wrong.

In case #3, parents gain little from offering complex new arguments, or even finding flaws in the usual kid arguments, at least when only parents can understand these. Parents instead gain from finding variations on the usual kid arguments that kids can understand, variations that get kids to do what parents want. Parents can also gain from talking at two levels at once, one discussion at a surface visible to kids, and another at a level visible only to other parents.

These three cases illustrate the three general cases, where your main audience is 1) about as capable , 2) more capable, or 3) less capable than you in generating and evaluating arguments on this topic. Your optimal argumentation strategy depends on in which of these cases you find yourself.

When your audience is about the same as you, you can most usefully “think for yourself”, in the sense that if an argument persuades you it will probably persuade your audience as well, at least if it uses popular premises. So you can be more comfortable in thinking sincerely, searching for arguments that will persuade you. You can be eager to find fault w/ arguments and criticize them, and to listen to such criticisms to see if they persuade you. And you can more trust the final consensus after your discussion.

The main exception here is where you tend to accept premises that are unpopular with your audience. In this case, you can either disconnect with that audience, not caring to try to persuade them, or you can focus less on sincerity and more on persuasion, seeking arguments that will convince them given their different premises.

When your audience is much more capable than you, then you can’t trust your own argument generation mechanism. You must instead mostly look to what persuades your superiors and try to parrot that. You may well fail if you try to adapt standard arguments to particular new situations, or if you try to evaluate detailed criticisms of those arguments. So you try to avoid such things. You instead seek generic positions that don’t depend as much on context, expressed in not entirely clear language that lets you decide at the last minute what exactly you meant.

When your audience is much less capable than you, then arguments that persuade you tend to be too complex to persuade them. So you must instead search for arguments that will persuade them, even if they seem wrong to you. That is, you must pander. You are less interested in rebuttals or flaws that are too complex to explain to your audience, though you are plenty interested in finding flaws that your audience can understand. You are also not interested in finding complex fixes and solutions to such flaws.

You must attend not only to the internal coherence of your arguments, but also to the many particular confusions and mistakes to which your audience is inclined. You must usually try arguments out to see how well they work on your audience. You may also gain by using extra layers of meaning to talk more indirectly to impress your more capable sub-audience.

What if, in addition to persuading best, you want to signal that you are more capable? To show that you are not less capable than your audience, you might go out of your way to show that you can sincerely, on the fly and without assistance, and without studying or practicing on your audience, construct new arguments that plausibly apply to your particular context, and identify flaws with new arguments offered by others. You’d be sincerely argumentative.

To suggest that you are more capable than your audience, you might instead show that you pay attention to the detailed mistakes and beliefs of your audience, and that you first try arguments out on them. You might try to show that you are able to find arguments by which you could persuade that audience of a wide range of conclusions, not just the conclusions you privately find the most believable. You might also show that you can simultaneously make persuasive arguments to your general audience, while also discreetly making impressive comments to a sub-audience that is much more capable. Sincerely “thinking for yourself” can look bad here.

In a world where people following the strategies I’ve outlined above, the quality of general opinion on each topic probably depends most strongly on something near the typical capability of the relevant audience that evaluates arguments on that topic. (I’d guess roughly the 80th percentile matters most on average.) The less capable mostly parrot up, and the more capable mostly pander down. Thus firms tend to be run in ways that makes sense to that rank employee or investor. Nations are run in ways that make sense to that rank citizen. Stories make sense to that rank reader/viewer. And so on. Competition between elites pandering down may on net improve opinion, as may selective parroting from below, though neither seems clear to me.

If we used better institutions for key decisions (e.g., prediction/ decision markets), then the audience that matters might become much more capable, to our general benefit. Alas that initial worse audience usually decides not to use better institutions. And in a world of ems typical audiences also become much more capable, to their benefit.

GD Star Rating
Tagged as: , ,
Trackback URL:
  • Brian Slesinsky

    An extreme case of “parroting” what experts say happens online, though the sharing of links. This is lower effort than writing things yourself, and often more useful, since the reader (who may know more than you) can go to the source and judge for themselves.

    As a sharer of links, it’s somewhat harder to make an embarrassing mistake; you’re judged mostly by the quality of your links, which means judging the argument’s plausibility, both to you and to your audience. Selectively quoting from the article tends to be pretty safe too. It can be difficult to do a better job at explanation than a skilled writer, and as you say, paraphrasing an argument risks introducing an error, particularly when moving quickly.

    In conversation, the equivalent might be to say “I don’t know, ask Google” or if you’re feeling generous, do the search yourself and have a conversation about what you find. In a conversation of which route to take, checking traffic on your phone will help you make a compelling argument.

    With link sharing, perhaps there are fewer opportunities for demonstrating intelligence, but more opportunities for showing good judgement, since you’re judged by the relevance and plausibility of what you share.

    The downsides are lack of originality and the likelihood of repeating misinformation when your judgement fails. But, for *most* subjects beyond those you have personal experience with (which is after all, most of what’s going on in the world), it’s probably a better bet.

    Perhaps younger generations will find most conversations without technical assistance to be ignorant, vague, and unreliable? The denigration of “mansplaining”, while mostly associated with feminism, might also reflect a growing disdain for speculative theorizing or philosophizing.

    But I expect there will still be plenty of places on the Internet open to such things.

  • Lord

    Somewhat overlooks other methods of dealing with arguments, changing the subject, casting doubt, privileged knowledge, us against them, avoiding them or making them beside the point.

  • You should have a name for that audience – eg “the reference audience”. If it’s right that the reference audience is substantially below the very best that could mean that the returns to greater skill diminish once you’ve passed the reference audience. The very best can come up with arguments which are too complex for the reference audience. Unless they restrain themselves from giving those arguments, that ability could have disadvantageous effects.

  • When your audience is much less capable than you, then arguments that persuade you tend to be too complex to persuade them. So you must instead search for arguments that will persuade them, even if they seem wrong to you. That is, you must pander.

    Must? You’re saying then that intellectual dishonesty is mandatory when addressing epistemic inferiors? Then you must pander (search for arguments you know lack merit) to your students?

    Your definition of pandering as searching for arguments you know lack merit is incisive, but do you really mean to say pandering is unavoidable? (The honest alternative, it seems to me, is actually a kind of parroting – an appeal to authorities.)

    • My strategy claims were relative to the goals I identified.

  • Jim Drogan

    Re “These three cases illustrate the three general cases, where your main audience is 1) about as capable , 2) more capable, or 3) less capable than you in generating and evaluating arguments on this topic. Your optimal argumentation strategy depends on in which of these cases you find yourself.”

    Suppose the audience and you have different opinions of capability. It seems to me that the optimal argumentation strategy and hence value of the conversation is is a function of the degree of agreement between the two perspectives.

    Furthermore, the strategy is adjusted by the relative power (i.e., power to compel an outcome) of the audience and you.

  • Pingback: Rational Feed – deluks917()

  • Pingback: Recomendaciones | intelib()

  • Richard Neumann

    Isn’t the problem that when you are more capable than your audience they usually parrot an “expert” opinion? If you try to convince them of otherwise it becomes too complex for them and they just rethrow the opinion at you again feeling they are right because you failed to convince them. This is one of the reasons whole institutions and norms in society can exist even though they are severely flawed.

  • davidmanheim

    Very interesting framing of classes of arguments. There is another dynamic I think might be worth identifying in this context, which is when a less-informed arguer attempts to present an argument to find why experts disagree, or where they are wrong.

    It’s especially important when it allows the people below that 80% threshold to get closer – a mechanism that internet conversations was once thought of facilitating. It turns out that the more common dynamic is… less inspiring.