Table of Contents
The standard view of debate uses weighted factors. Arguments are factors which add support (or strength, weight, points, justification, etc.) for a side. Critical arguments subtract instead of adding. Arguments have different weights which determine how much they add or subtract (some arguments are stronger than others). A debate is judged by which side is the strongest/best after adding and subtracting all the arguments.
Critical Fallibilism (CF) says the weighted factor approach is mathematically and logically invalid – it doesn’t and can’t work. CF offers an alternative approach focused on decisive arguments. CF focuses on binary judgments (like pass/fail) for arguments instead of judging degrees, strengths, weights or amounts of goodness or badness. CF uses a digital approach for judging ideas instead of an analog spectrum. CF also focuses attention on a few key issues instead of adding up every minor issue or factor. CF also leaves qualitatively different factors separate instead of trying to combine them.
Private Relay Example
The MacRumors article UK Network Operators Target iCloud Private Relay in Complaint to Regulator reports on biased, one-sided arguments against Private Relay (a privacy feature from Apple). The group opposing it make no attempt to give a fair, comprehensive or neutral summary of the issues. They just say what’s (allegedly) bad about Private Relay but don’t attempt to say what’s good about it. They don’t discuss its intended purpose (what problem is it trying to solve?) or offer an alternative way to achieve the goal of Private Relay (nor do they explain why that goal – privacy – is bad). They argue against Private Relay because they believe they have an incentive to oppose it. They think it’s in their interests to get it banned or limited so that they can make more money off user tracking that enables targeted advertisements.
Apple and other proponents of Private Relay focus on its upsides like privacy without giving much attention to the upsides of advertising. They, too, present a biased case.
Arguing for “your side” of an issue – focusing on the arguments that help your case but not giving arguments against yourself – is considered normal in our society. But it’s biased, not truth-seeking. It’s the kind of thing scientists aren’t supposed to do (but often do anyway).
The underlying model, for how people look at debate, is weighted factors. People use indecisive arguments to add weight (or strength or impact) to their side. They may also attack the other side(s) to remove weight. But they don’t want to add any weight to the other side or lose any weight from their side, so they won’t say any arguments for the other side or against their own side. The assumption is that opposing people will do the same thing, trying to add weight only to their side. Then the audience will favor the stronger side.
Critical Fallibilism (CF) offers a different model. Instead of incentivizing biased one-sidedness, CF’s model incentivizes engaging with the other side’s points.
The weighted factor model results in both sides basically independently, separately making their own cases, and then the cases are compared by total strength. There’s little interaction or engagement between the sides. There’s limited questioning or grilling because people would rather argue for their conclusion than participate in the other side’s case by responding to challenging questions. There’s limited counter-arguing because both sides would rather focus on the positives for their side without drawing attention to negatives (counter-arguments draw attention to the criticism they’re trying to counter). Giving counter-arguments is seen as a sign of weakness – you only do it if the criticism is so strong and threatening that you have to try to answer it.
CF says to instead look for key, decisive issues. In short, find an issue that could be decisive, then challenge the other side with it and await a response. There’s no harm in talking about arguments and ideas for the other side as long as they aren’t decisive. If you think you’re right, then you think the other side doesn’t have decisive arguments, so you shouldn’t be scared of their arguments. You shouldn’t mind answering questions and responding to arguments, and talking about why their side isn’t decisive. You have to talk about their arguments in order to point out that they aren’t decisive.
CF says to view debate as a tree with responses and counter-responses, and to try to reach a conclusion by evaluating the whole tree. And there’s just one tree, not a separate tree for each side to make its own independent case.
Instead of looking at the total weight of arguments for each side, you should look for unanswered decisive arguments. The weighted factor model means that you should use every single bit of audience attention you can get to add more weight to your side (or reduce the weight of their side). But the decisive argument model means a adding a few more low-quality, non-decisive arguments to your side doesn’t help. That’s not worth any points or weight.
So, once you make your main points, you can spend your remaining time (that you can get the audience to pay attention to you) clarifying the tree of the debate, organizing the arguments, giving rebuttals to erroneous arguments, and summarizing the state of the debate. You could also keep repeating or clarifying your key, decisive arguments, but at least there are only a few of them, and there’s no incentive to keep piling on more arguments of varying quality or importance. And there is an incentive to engage with the other side, in order to explain why their allegedly-decisive arguments don’t work, and to defend your decisive arguments from refutations.
Instead of making sure only to talk about things that are good for your side
If people thought with CF methods then it’d be OK to give attention to the other side’s arguments because what matters is analyzing what’s decisive and why. You wouldn’t have to stick to only saying things that are good for your side, because the goal would be to figure out how to reach an objective conclusion. Instead of giving biased arguments, you’d be expected to help explain to the audience how to reach a neutral, objective, fair conclusion.
Also, in the weighted factor model, in practice, perceived strength of arguments and attention arguments get are closely related. People tend to think that whatever they heard the most and/or most recently seems the most persuasive or strong. However, if the focus is on judging whether an argument is decisive or indecisive, then that requires the audience to look at it in a different way, that isn’t a matter of degree, so it’s harder for them to actually judge by the wrong factors like familiarity (which is a matter of degree).
A decisive argument is a reason that something must or cannot work (at a purpose, objective or goal). The something could be an idea, action, or plan. Actions and plans are actually types of ideas. We act according to ideas. We have ideas about what actions to take. All debate is, in some way, about ideas.
A decisive argument contradicts at least one alternative. It’s incompatible with something. The argument, and the alternative, cannot both be correct (as best we understand it – it is possible for a new argument to point out that we were wrong about what contradicts). You must choose one or the other (or neither). This is different than strength-based arguments because you shouldn’t conclude that the argument is pretty good but the thing it tries to refute still seems pretty good too. They contradict. At least one is wrong. They aren’t both pretty good. If you conclude that they’re both pretty good, you are guaranteed to be wrong. Your conclusion has a logical contradiction in it. If they both seem pretty good, it means you don’t understand the issues and should instead conclude that you don’t know.
Decisive positive arguments run into difficulties with fallibilism and with the difficulty of proving a positive claim. How can you rule out all logically possible alternatives in order to prove a positive claim? Generally, you can’t.
So CF focuses on decisive, negative arguments. We can settle debates only using those. Positive arguments can be OK as an approximation or shortcut if you know how to convert them into negative arguments if asked to. (Many positive arguments that people use are equivalent to negative arguments, so they’re actually OK, just a little sloppy. We don’t need to reject them all. But when being precise and rigorous – which isn’t always necessary – we should convert them to negative arguments. Any positive argument with no equivalent negative version is mistaken.)
So instead of arguing about how great your side is, you have to present potential or actual rival ideas and criticize those. In order to use decisive, negative arguments, you have to talk about ideas that aren’t on your side (otherwise you’d be refuting your own side).
CF’s model of debate has two basic parts: you explain your side and argue against alternatives. Explaining what your idea/side is and says, and how it works, is different than argument. Arguments are meta-ideas about some other idea, which argue why it’s correct, incorrect, good or bad. So first people can share ideas/solutions/proposals and then second they can argue about those ideas. The first part is positive but it isn’t argument, and the second part is negative argument. It’s like how stating your thesis and arguing for it are different things.
The US court system is another example where weighted factor arguing is used. It’s an adversarial system with a lawyer (or team) on each side (and a neutral judge, who has parallels to a referee in sports). Each lawyer is expected to make the best case they can for their side. They aren’t expected to be unbiased or fair. It’s their job to argue for only their side. They aren’t supposed to say anything that harms their case, which would be seen as betraying their client. (Some other countries have different legal systems. All of them have some problems.)
In court, there’s basically is no punishment for giving weak arguments (besides sometimes having a time limit, so they need to prioritize their better arguments before time expires). Lawyers often raise dumb objections and even feel pressured or required to do that, despite knowing their claim is dumb and will probably be rejected by the judge. Why? Because it’s their job to do everything they can to advocate for their client/side. A weak argument is seen as better than no argument. They have to put up a fight and try.
In a rational system, people would be expected to have integrity and try to make only arguments they think are actually pretty good and worthwhile. One way to view that is you should make decisive arguments, rather than arguments which are compatible with you losing the case. In a good system, each side would try to give a complete picture of the issues that explains the right conclusion to reach, rather than to pay selective attention to why their side is so great.
A rational case, even if you’re biased, involves trying to explain the important arguments on both sides, and then how to objectively reach a conclusion which takes into account and addresses those arguments. It looks for key, decisive factors. It points out what doesn’t matter, where attention should be focused, and why. It explains a conclusion which is not merely stronger than an alternative but non-refuted while the alternative is refuted. (Refutation is contextual, so people should also specify some goals or standards. You need some goalposts, and to argue about why the opposing case fails to reach the goalposts. You could never successfully argue that it would fail to reach all possible goalposts.)
The weighted factor model encourages both sides to talk past each other. They list lots of pros for their side and some cons for the other side. The focus is usually more on pros because they don’t want to give attention to the other side. And reasons your thing is great are considered better than sniping at the other guy’s thing. Attacking the other side can also be seen as mean. And you can’t win just by playing defense. And critiquing the other side rarely gets their score below zero. You win debates mostly by getting a high positive score, not by leaving your side’s score low and trying to get the other side’s score even lower with criticism.
Note: In practice, I don’t think people consistently stick to the weighted factor model. Sometimes they make some rational arguments about decisive issues.
The CF model says you win debates by organizing the issues in a tree (or using another organizational method), showing the big picture, and explaining why you have a decisive conclusion. Your conclusion should be a win/win solution – it should get a good outcome for every goal you care about (if you can’t do that, you need to reconsider your goals rather than keep the same goals and fail). You should show that your conclusion is non-refuted by engaging with attempted refutations.
People should put forward a few key claims and explanations, then critique the other side. They should discuss and argue about what the overall, objective state of the debate is. People should discuss which arguments are decisive or indecisive (in other words, what contradicts what, and what is non-contradictory). This is more rational than each side separately trying to build up their position with limited engagement with the other side.