Inverting Paul Graham’s “How to Disagree”: “How to Assert”
In Beyond Good and Evil, aphorism 48, Nietzsche wrote a striking passage in which he quotes something he disagrees with strongly to the point of almost exact opposition, but then, instead of responding with indignation against it, he follows with evocative gratitude: “how neat, how distinguished to have one’s own antipodes!”
I admired this in my early twenties and have practiced it ever since (though often imperfectly), so this experience of feeling grateful to the world for containing one’s own exact opposites has become a habit for me. I think it’s a very natural sense of things if you look at the intellectual world as more of an ecosystem than a tournament bracket: the natural world is filled with complementary symbiotic opposites, from cleaner wrasse and grouper to squirrels and oaks. Intellectually, I find I deeply rely on my opposites for symbiosis as well.
Therefore, since I value opposites and complements, I often multiply them. I take whatever saying or essay I’m presented and imagine a lattice of different inversions of the material, sort of like flipping sequences of different bits of the central message of the essay while leaving the surface structures unchanged. (Programmers and mathematicians may recognize “homomorphism” intuition there.)
Given an essay celebrating plants, I might immediately imagine rewriting the essay in three ways with the same general structure: celebrating rocks, excoriating plants, and excoriating rocks. Seeing exactly what must change and what must not to get a precisely but dramatically changed meaning is most of the fun.
Usually I leave these inversions unspoken and unwritten and they just fill shelves of an inner imaginary Library of Babel for me, but if I’m going to be sharing more of myself online I think these are some of the most natural examples to start with. I hope the inverted essays themselves are thought-provoking, but even more I hope this pattern of grateful, playful multiplicity comes across through them. Because really — isn’t it neat and distinguished to have one’s own antipodes?
The first of these inversions I’ll share is a play off of Paul Graham’s perennial “How to Disagree.” I consider the essay good in itself and in its niche, but also often exasperating to me in my niche and opposed to my spirit in my niche. In any freeform creative discussion group where it becomes an authority, I expect to lose interest in that group’s discussions within a few months, because while the essay is effectively correct on its own terms, I see it have an unbalanced influence on social patterns. It turns these groups grayer, and duller. It’s like a pesticide that, while killing nuisance insects, also throws off the balance of the soil and the taste of the crop. Where routinized, industrious conversation is the order of the day, it may be perfectly in order. Where more organic conversation is desired, it’s a killer.
So I thought up a defense of name-calling, ad hominem, and responding to tone: one a little more serious than just calling it “chad hominem,” but still not a straight, rigorous defense that would just pick a fight and risk reproducing the same patterns of grayness and dullness itself. Instead, I inverted “How to Disagree” with three flips: I flipped the hierarchy, I made it about assertion instead of disagreement, and I centered subjective actual-audience judgment over objective hypothetical-audience clarity. This last flip is most important to me, because the central problem I see with “How to Disagree” in context, the one that does the most to make it an unbalanced social influence in practice, is that most of its readers dramatically overestimate how objective and self-evident their arguments and refutations are.
The point here isn’t to make the serious defense of name-calling, ad hominem, and responding to tone (i.e., of drawing attention to honor, ethos, and pathos, in terms of the Classical rhetorical tradition; it could be done), but just to play around enough with promoting them that the apparent obviousness of condemning them becomes harder to take for granted.
“How to Assert” is not a strong essay stand-alone, and it’s not meant to be. It’s a game. It might inspire a strong essay later, but probably not one I write. It’s also not meant to destroy or refute “How to Disagree.” That essay is deservedly perennial in my opinion, and the goal here is to balance it, not to break it: it’s to give it one of its own antipodes. Healthier squirrels and healthier oaks go together, and I hope “How to Assert” and “How to Disagree” will complement one another from here on as well.
I recommend reading the two essays alongside one another to see the similarities and differences exactly. I followed the original closely. Graham’s essay is here: https://www.paulgraham.com/disagree.html.
How to Assert
April 2022
Gaul Landau
Social media is turning conversations into trials. Fourteen years ago, conversationalists conversed and participants participated. Social media lets non-participants police, and increasingly they do — in boycotts, in deplatformings, and in cancellations.
Many who police something bond with each other through opposing it. That’s to be expected. The negation of any specific, substantive claim is more general, thus more prima facie plausible and more likely to attract support. And when you oppose something together there’s a grounds to start bonding. You could bond with the author making substantive claims, but he’s probably going to want more specific commitments. When you bond through agreement about opposing and policing others, you don’t have to prematurely commit to each others’ positive programs.
The result is there are a lot more policing coalitions around, especially measured by strength of their vetos. This doesn’t mean people are getting more friendly or considerate. The structural change in the way we communicate is enough to account for it. But though it’s not consideration that’s driving the increase in policing, the increase in policing does cause would-be writers to consider their words more carefully. Particularly online, where it’s easy to speak to audiences with unanticipated sensitivities.
If we’re all going to be policed more on substance, we should be careful to bear it well. What does it mean to assert substance well? Most participants can tell some difference between speech that is merely making a point and speech that earns trust, but I think it would help to put names on the intermediate stages. So here’s an attempt at an assertion hierarchy:
AH0. Forcing a Point.
This is the lowest form of seeking agreement, and probably the most common. We’ve all seen out of context posts like this:
Philosophical realism is trash.
But it’s important to realize that more articulate point-forcing has just as little weight. An out of context assertion like
The only philosophy that properly integrates the empirical with the a priori is German Idealism.
is really nothing more than a pretentious version of “philosophical realism is trash.”
AH1. Forcing a Premise.
Forcing one’s premises is not quite as weak as merely forcing one’s point. It might actually change someone’s mind. For example, if a senator anticipated receiving pushback for asserting directly that his salary should be increased, he could assert:
American senators are underpaid.
This wouldn’t necessarily convince anyone, but it might at least be relevant to why the audience oppose paying him more. It’s still a weak form of assertion, though. If there’s opposition to your points, you should anticipate why that is; and if you haven’t, what difference is a small, transparent indirection going to make?
Insisting that others need to read the books one has read to oppose an assertion is a variant of forcing one’s premises — and a particularly useless sort, because so many books are fraudulent or otherwise low-quality. The question is whether the assertion is convincing or not. If the books actually make a strong case for the point, make that case yourself. And if the books don’t, no one owes it to you to prove that before rejecting the thin case you did make.
AH2. Forcing a Frame.
The next level up we start seeing a semblance of engagement with the audience as persons, rather than mere assertions of our own points as if they had a self-evident logic. The lowest form of these is to implicitly bully the audience into conformity with the patterns of reasoning used to establish the central point. E.g.,
Bayesianism is the only formal epistemology that doesn’t suffer a Dutch Book problem [and so implicitly anyone who is not a Bayesian loses their bets and implicitly such persons are losers].
Though better than just asserting premises, this is still a weak form of assertion. It matters much more whether the audience is made better off by listening than just whether they change their beliefs or not. Especially since conformity-driven beliefs can so easily change again. A changeable person can be bullied a dozen times in a day, so the less changeable ones can be far more valuable to persuade.
So if the best assertion you can muster is an intimidating frame control, you’re not going to get far. Is the audience firmly committed to an opposed frame, but on definite principles? Better that than changeable and unprincipled. If your audience’s frames for your point are mistaken, engage with why.
AH3. Parroting Premises.
In this stage we finally see engagements of others’ prior judgments about a point, rather than pretenses that the point should be a self-evident fact. The lowest form of these is simply to assert that your point follows from others’ stated premises by repeating those premises, with little or no checking that their words have the substantive semantic implications you assert they do.
This is often combined with AH2 frame control, as in:
I understand that you think Communism is a good idea. But you also believe that people should have good lives. And people never have good lives under Communism; the only people who say otherwise are Holodomor-deniers [and you don’t deny the Holodomor, do you?].
Parroting premises can sometimes result in a persuasive argument. Sometimes merely having your premises parroted back at you does spur self-reflection and a change in views. But usually meaningful engagement will help.
AH4. Convincing Tone.
At level 4 we reach the first form of persuasive assertion: presentation with convincing tone. Forms up to this point will usually be dismissed as out of touch. Something with a convincing tone might actually connect with an audience. The problem is, it’s hard to check for substance.
Establishing a convincing tone is repeating others’ premises plus at least a semblance of similar feeling for the spirit of the premises. But unfortunately it’s common for those semblances of common feeling to be false. More often than not, someone speaking persuasively still does not actually have their audience’s best interests in mind. Sometimes they themselves don’t even realize this, caught up in their own pretense of goodwill.
There can be legitimate reasons for adopting a convincing tone even when you are deceiving another: when the deception is a white lie necessary to keep the peace, for instance. But when you do that, you should make sure no one gets hurt through believing you.
AH5. Credible Case.
The most persuasive form of assertion builds a credible case for the central point beginning from the audience’s premises and in the spirit of those premises. It’s also the rarest, because it’s the most work. Indeed, the assertion hierarchy forms a kind of pyramid, in the sense that the higher you go the fewer instances you find.
To be credible to someone you probably have to understand them. You have to find their real concerns and then provide concrete evidence allaying those concerns that they can check for themselves. They may have justified mistrust of you. If you can’t find those real concerns that guide your audience’s judgments of your point, you may be trying to exert control rather than actually persuade.
While credibility generally entails understanding, understanding doesn’t necessarily imply credibility. Some authors use their understanding to find an audience’s personal vulnerabilities, then follow with a tactic as low as AH3 or even AH0.
AH6. Earning Trust.
The virtue of a credible case depends on what you make the case for. The best credible cases are those that earn lasting trust from the audience.
Even as high as AH5 we still sometimes see barely-veiled manipulation, as when someone makes a credible but unwise case to convince an audience to make a poor decision. Sometimes the spirit in which this is done makes it more of a sophisticated forcing of premises than an actual credible case. For example, appealing to a wrongly-trusted authority, or insisting that a matter is purely rational when it has subtle emotional dimensions. Unless the audience’s true concerns actually are addressed by these premises, the only purpose of asserting them is to bully one’s audience into agreeing to things they will quickly regret.
Truly earning the trust of an audience requires accountability for how your persuasion impacts their concerns and whether the credibility of the case survives application. And that means you have to be willing to stake your good name on your writing. Some comments you get if you are successful will look like:
I was not sure whether to trust you, because you are <name-calling, ad hominems>. But you made a good case, you convinced me, and I’m glad you did. I’ll vouch for you in the future with my side, though please keep up your standards and don’t start taking advantage. You’re really not what I expected for a <name-calling, ad hominems>. Thanks for helping me out.
The name-calling and ad hominems might not seem the least bit reasonable to you. Still, they represent the baseline mistrust that you have to overcome to make your own name with an expanding audience.
What It Means
Now we have a way of classifying forms of assertion. What good is it? One thing the assertion hierarchy doesn’t give us is a way of picking a winner. AH levels merely describe the appeal of an assertion to an immediate audience, not whether it’s acceptable in general. An AH6 proposal could still be transparently manipulative to bystanders who then come in to police it.
But while AH levels don’t set a lower bound on the persuasiveness of an assertion, they do set an upper bound. An AH6 proposal might be unconvincing, but an AH2 or lower proposal is always unconvincing.
The most obvious advantage of classifying the forms of assertion is that it will help people to evaluate what they read. In particular, it will help them see through shallow, poorly motivated bids for persuasion. A sophisticated speaker or writer can give the impression of authority before an audience merely by using expert-sounding words. In fact that is probably the defining quality of a charlatan. By giving names to the different forms of assertion, we give critical writers a pin for popping such balloons.
Such labels may help writers, too. Most intellectual dishonesty is unintentional. Someone forcing a frame he believes in may believe he’s really helping his audience. Zooming out and seeing his current position on the agreement hierarchy may inspire him to try moving up to using a convincing tone or making a credible case.
But the greatest benefit of asserting well is not just that it will make conversations better, but that it will make the people who have them happier. If you study conversations, you find there is a lot more insecurity down in AH1 than up in AH6. You don’t have to be insecure when you have real value to offer others. In fact, you don’t want to. If you have something real to offer, insecurity just gets in the way.
If moving up the assertion hierarchy makes people less insecure, that will make most of them happier. Most people don’t really enjoy being insecure; they accept it because they can’t help it.
Thanks to Rover Whitetower and Leshiqua Jivingston for reading drafts of this.
All Said
It’s a long, dry sort of exercise if you actually spell it out all the way like this, and I don’t do it often (this one is from a few years ago). It’s like a heavy homework problem chosen in order to strengthen a faster and lighter thought pattern. It’s deliberate, plodding practice meant to enable more fluid imagination and snappier argument later.
With practice, it becomes an aesthetic pleasure: I’ve come to love the lightning flash of intuiting a lattice of 8 combinations of three single-bit inversions in my head, then feeling the slower, rolling thunder of growing inferences about how each combination could imply different social and moral impacts. But that may or may not be credible to you if you’ve never experienced it, and I won’t force the point!
I do believe the substance of what I wrote in “How to Assert” and I hope the general theme has resonated: good conversation requires respecting that others don’t need to take propositions on faith or charitably outside of pre-selected discussion groups with strong prior norms to that effect. An ad hominem is never a refutation of an argument, but it can efficiently and correctly deny standing to a bad faith argument from someone who never deserved the charity required to properly refute it. Not an original point, granted, but it’s one worth perennially repeating in new ways.