Argument by Intellectual Ordeal

post by lc · 2022-08-12T13:03:21.809Z · LW · GW · 5 comments

There once was an extremely annoying child named Jacob. One time during recess, Jacob decided to take up the entire playground bench by lying down on it horizontally. A classmate came up to him and said something like: "You're taking up the whole bench! That's not fair!" Jacob responded: "Ugh, you people don't know what I'm doing! Do you even know what quarks are?!" The classmate said "umm..." thoughtfully for a few seconds, and then scuttled off.

While an effective rhetorical technique, knowledge of the existence of quarks was not a very relevant indicator for who was right. The question didn't test for knowledge relevant to playground bench allocation problems, didn't correlate well with even general thinkoomph, and neither of these would've correlated well with correctness in the first place because of Jacob's clear conflicts of interest.

Medieval Europe had a method of determining guilt called "trial by ordeal". If two parties had a large dispute - say one accusing another of a serious crime - they would sometimes decide who was right by fighting to the death in an arena, or having the accused willingly ingest poison and see if they threw it up instead of getting sick or dying. The presumption was that God would favor the one telling the truth, or maybe that their skill would be bolstered by their sense of righteousness.

I like to pejoratively call the type of fallacious impromptu challenge-giving I just described "argument by intellectual ordeal". The lifecycle of the rhetorical tactic goes like this:

1. Person A makes an argument: "You taking up the entire bench when multiple people could sit on it seems unfair."
2. Person B counters with a mostly irrelevant challenge testing Person A's knowledge, general intelligence, or other: "If you're so smart solve 18283921 x 192399213 in the next ten seconds!"
3. Person A, caught up in the emotional fervor of arguing with B or their own competitive instinct, attempts to solve the problem and fails.
4. Having lost the impromptu contest, which they legitimized by being observed attempting to win, Person A meekly concedes the point.

Like "arguments from authority", the "argument by intellectual ordeal" comes with standard justifications that are almost always overused. In the case of the "argument from authority", the surface justification is something like "People with the X rubber stamp tend to be very good at answering these questions, so it's only natural that we defer to their expertise even if we think we have a gears-level model that says they're wrong." Much of Inadequate Equilibria [? · GW] is a long case study as to how, on deeper inspection, appeals to expertise can be a smokescreen for a more pernicious and instinctual aversion to Intellectual Immodesty, and not really predicated on that authority's incentives & track record.

When it comes to "argument from intellectual ordeal", the surface justification given by proposers is usually that succeeding at the challenge would either suggest some domain-specific competence or knowledge, or highlight the consistency & predictive power of the other person's world model. The pernicious underlying rationale, is that when someone is unable to respond to any surface-relevant challenge or inquiry raised by interlocutors, that seems enough to make one 'lose hit points' in a debate, whether or not the challenge tested something meaningful in the first place. 

I think most people would agree that the examples I've shown so far are silly, and it'd be a social faux pas to make the demands to adults. Most real world examples of argument by intellectual ordeal however aren't quite so clear cut, because they're either more directly relevant to the topic or are embedded in more innocuous lines of questioning. Often it's just the way a critique is phrased into a question that turns it into a challenge and not an argument. Let me show you some examples I've been involved in or witnessed, and explain why I think there were better alternatives. Note that I'm not taking a stand here on the debates, just noting the tactic as present:

• Person A: "I've tried drinking tons of BANG energy drinks after blacking out and it's basically solved my hangovers. You need to try it."
  Person B: "Ok, can you explain how it would stop acetaldehyde accumulation?"
  • Surface justification: Acetaldehyde accumulation is an important part of hangovers and if this person can't explain how BANG energy drinks solve that they're obviously wrong.
  • Problem: Whether or not this person understands what "acetaldehyde" is, they're not claiming their conclusion is based on a mechanical understanding of hangovers, they're justifying it from anecdotal experience. The fact that they in particular can or can't make up a story about how it solves this one aspect on the spot is not really relevant to their case. If Person B thinks this is relevant they should start by stating so instead of prompting for an explanation.
  • Less confrontational alternative: "I don't see how caffeine could prevent hangovers, because it doesn't affect acetaldehyde accumulation. Here's why that's important..."
• Startup founder: "We've been working on this product..." [Begins basic pitch and growth analysis]
  Potential Investor (interjecting): "So what's your business going to look like six months, one year, two years from now?"
  • Surface justification: It's important that these founders have a plan for what their long term future looks like and have a plausible long term vision that makes me money.
  • (One) Problem: The question is asking quickly about some very specific projections, and so it's mostly testing the founder's ability to come up with three nicely connected descriptions of the future on the spot, not just that they have a good plan persay. Most founders aren't going to have committed projections of those specific dates to memory even if they do have a coherent business plan. In addition, the fact that the person's pitch was interrupted means that they've lost their train of thought in a high pressure situation and are going to need to mentally "unwind the call stack". The stakes and manner of questioning tampers with the experiment even if testing their intelligence was your real goal.^[1]
  • Less confrontational alternative: (After waiting for their pitch to finish or hit a lull) "So, tell me how your business is going to evolve over time." [+ followup questions]
• One hacker: "I would never buy a Tesla. Some brand of cars is going to watch their internet connected engines all stop at the same time one day."
  Another hacker who previously expressed interest in the cybertruck: "Can you tell me how that would even happen? These self aggrandizing computer security horror stories never really seem to come to fruition." 
  • Surface justification: Obviously a sufficiently granular explanation from Hacker A of how all Tesla Roadsters could be stopped at the same time would lend great credence to his argument.
  • Problem: The original person is probably relying on general intuitions about DoS attacks and previous Tesla security scares they read about on hacker blogs, and isn't in possession of like, working attack plans.
  • Less confrontational alternative: Depends on the person's real objection. Perhaps: "My intuition is that these kinds of terrifying stories don't usually pan out unless they're based on technicals. Most of the large tech companies don't let obvious catastrophic risks like that remain unsolved for a long time unless they want to." 

Sometimes, very rarely, appeals to authority guided by a correct notion of market efficiency can be useful. I don't think I've ever encountered a situation where "argument by intellectual ordeal" was the best available option, in practice, except when the person's aptitudes were the explicitly acknowledged and unanimously accepted topic of conversation. So I have two pieces of advice for potential interlocutors. First, if it seems like someone has "challenged" you during an argument, don't just lean into the challenge reflexively unless you're absolutely sure you can win it; take a few extra seconds to make sure the challenge is relevant and then state your objections if you have them. If you do attempt the challenge and fail, it will be harder for you to claim later that the exercise was meaningless even if that's true.

Second, try to keep conversations conversational. Ask people questions primarily to understand their ideas and pass their Intellectual Turing Test, not to win a game of anticipating some hard-to-see consequence of their map (unless respectfully examining their ability to do that is actually your primary goal). If you have objections, state those explicitly and then give examples of arguments that would change your beliefs, instead of unilaterally demanding responses of a certain kind.

1. ^{^}

   In hindsight, the fact that the question was delivered this adversarially may have been deliberate. As I later realized while reading Elder Rationalist Zvi's posts, one alternative explanation is that it was testing presentational skills useful for impressing pseudo-incompetent downstream investors. In this case the question might have been much more plausibly relevant than it would first appear.

5 comments

Comments sorted by top scores.