Table of Contents
Critical Fallibilism (CF) loosely separates debate into two parts.
First, you explain your idea. You present it and say what it is and how it works. This is not arguing how great or strong it is, nor arguing that your idea is correct. Sharing ideas is different than arguing. You may address some potential questions. What goals does the idea achieve, and how does it do that? Why does it make sense? What earlier ideas is it based on if any? Is it a modification of a prior idea to solve some problem? What is the history of the issue and why did older ideas fail and how does your idea do better?
Second is criticism. This is where you argue. If you present an idea and also refute all criticism of your idea, then your idea is non-refuted. If you give decisive criticism of alternative ideas (competitors, rivals), then they’re refuted. If there is exactly one non-refuted idea, then it wins the debate.
Note: You have to start debates with the first part (presenting an idea). After starting the second part (criticism), you can still go back and do more of the first part. But you can’t criticize first, before there’s anything to criticize.
The Logic of Debate
If we have multiple non-refuted ideas which contradict each other, and we can’t resolve the matter, then we should reject all of them. One may happen to be correct, but we don’t know it; none of them are actionable knowledge. Each one failed to refute an alternative that contradicts it, so it’s wrong to reach its conclusion. If you can’t refute X nor refute not-X, then reaching the conclusion X is an error. Reaching the conclusion not-X would also be an error. You should conclude that you don’t know. Your knowledge is insufficient to choose between X and not-X, and picking one over the other is arbitrary and biased.
If we have multiple non-refuted ideas which don’t contradict each other, then we can either use any of them (there are multiple viable solutions) or consider changing to a more ambitious goal. If we have no non-refuted ideas, we can either brainstorm more ideas or consider less ambitious goals.
It doesn’t matter how “strong” your initial case is when you propose an idea. You just have to have an idea which addresses a problem (or goal, objective or purpose) and communicate it. If you do that then, barring criticism (a reason it wouldn’t work as intended), your idea is a winner. If your idea is weak then it’ll be easy to criticize. But instead of trying to judge whether the idea is strong or weak, good or bad, you should look at the actual results of critical thinking. What criticism of it do we come up with? Instead of trying to predict how an idea would do in critical debate (strong means it would do well in critical debate, and weak means poorly), we should have the actual debate and look at the results. (You can quickly have a version of that debate in your head if you’re capable of being neutral enough to think of arguments on both sides. Debating with other people takes longer, and can be reserved for difficult, important and controversial cases, not every little thing.)
Every criticism (decisive refutation) of an idea must be refuted or else the idea is refuted. It can be useful to divide refutation of criticism (counter-criticism) into two types. First (standard type), you can point out a flaw (error, mistake) in a criticism. Second, you can say how a criticism is compatible with its target succeeding at its goal(s). In other words, even if the criticism is correct, the idea it criticizes still isn’t refuted. The criticism is indecisive. The criticism doesn’t contradict its target.
Many alleged criticisms are not really criticisms because they are compatible with the thing they criticize. A proper criticism contradicts an idea, or in other words says it will fail (at a purpose/goal/problem/objective). Any criticism that merely says “it’s weak” or “it’s not great” is not giving any reason it will fail at any particular thing, and is therefore a bad argument. The thing that matters is success or failure, and for a criticism to be relevant and be a potential refutation, it has to actually allege the idea will fail at something. A criticism has to connect something about an idea to failure at some goal; otherwise it’s either vague or wrong. And, to matter, a criticism needs to be about a goal we actually have or at least a candidate goal we’re considering having.
And when ideas are refuted, we have to keep track of which goals they’re refuted for. Instead of concluding “this idea has high strength but some flaws” we should conclude “this idea is decisively refuted for goals X and Y, but is non-refuted for goals Z and W”. The CF view is more actionable: you can still use the idea for Z but should not use it for X. If you know only that the idea has a high, low or medium overall score, then you might use it for any goal and have no idea that its score should be zero for some goals and infinity for others.
Another View of Argument Strengths
Another way to think about argument strengths is that explaining the whole state of the debate, and why you think it reaches a particular conclusion, is a really strong argument. It’s the best argument you could make. It’s a conclusive, complete argument instead of a partial argument. Decisive arguments are stronger than arguing about individual factors in different dimensions that probably aren’t bottlenecks anyway. You only can know what’s a bottleneck by looking at the big picture because a limitation of one factor in isolation may or may not actually constrain your success at your overall goal(s).
In functional systems, most factors have more than enough and we shouldn’t put effort into increasing or optimizing them. In terms of debate, most possible arguments about how great something is are trying to increase the strength of a factor that’s already good enough and doesn’t need to be better. Making arguments to try to strengthen random things on your side is just like going to a factory and trying to improve the productivity of random workstations. That is mostly wasted effort. Instead, we should focus our arguments on a small number of key issues (like factories should search for bottlenecks and focus their improvements there).
It’s not really that decisive arguments are strong. It’s a qualitative difference: they are complete arguments that enable reaching a conclusion, as against local isolated arguments about individual factors. Those are incomplete or partial arguments – they are building blocks that could be used in an overall view of the matter, but they don’t actually reach a conclusion.
A third alternative to complete or partial argument is no argument. These are not degrees of strength; they are three categorically different things that should be handled in different ways. They are three discrete things, not a spectrum. You can have reasoning adequate to reach a conclusion, reasoning that could be helpful but is not currently able to reach a conclusion, or no reasoning.
People look at argument strengths because they don’t know how to think decisively or organize a whole debate. Disorganization leaves people stuck seeing a bunch of local factors but not a whole debate tree that reaches a conclusion. So they try to add up the different-dimension factors because they need an overall judgment, even though some factors are not bottlenecks and the list of factors they consider is incomplete. It’s also hard because many factors aren’t numeric (e.g. how good looking a car or piece of clothing is). And the weights and evaluations people give to factors are pretty arbitrary due to the mathematical impossibility of combining numbers from different dimensions with addition.
People also look at argument strengths because they feel intuitions of different strengths. It’s hard to figure out what to do with your intuitions in a rational, organized, explicit debate. People often struggle to turn intuitions into words and reasons. With weighted factors, they can give their intuition as high a weighting as they want and let it compete on even footing with factors that are actually stated in words.
Some would see this as biased, irrational clinging to intuition. I don’t. Intuitions contain knowledge. They need to be addressed, not ignored or suppressed. Just discounting or ignoring intuition because it doesn’t offer English arguments would be wrong. But don’t just add weight to conclusions based on your intuition. Intuition can be handled in other ways. Here are two ideas for what to do when you think an issue is worth putting effort into and you have an intuition about it that feels important:
- Explore the intuition. Figure out what the concern, goal, criticism or issue is. Putting the intuition into words is one option. But you can learn a lot by asking yourself questions and seeing how you intuitively feel about different scenarios.
- Ask “Given I have this intuition (and given various other things I know), what action should I take?” If you don’t put your intuition into words for discussion, then you need to find a way of proceeding which is compatible with the intuition’s existence. Your explicit reasoning can take it into account.
You can either understand an intuition (and bring it into rational, explicit discussion) or treat it as a given – a part of the situation – and rationally, explicitly discuss what to do about that situation. Examples of other givens include “I have only $200 in my bank account”, “I’m short”, “I don’t know how to repair a car” or “Joe and Sue don’t get along”. These are facts of the situation which your ideas should handle appropriately. E.g. don’t plan to repair a car yourself or to seat Joe next to Sue. However, as always, it’s contextual. E.g. if you’re a reality TV producer then you might seat Joe next to Sue on purpose because you want some drama.
What should be done given that you have certain intuitions that you don’t understand well? That is a thing you can rationally discuss without needing to put those intuitions into words. You can rationally figure out what to do, using conscious thought in a language like English, without ignoring your intuitions, but also without being stuck due to the inability of those intuitions to participate in that discussion.
How do you explore an intuition without getting it into words or conscious reasoning? Ask yourself if we make this change to an idea, does that intuitively seem better or worse to me? If we do it this other way, does my intuition stop objecting? You can keep proposing things and seeing what your intuition says about them until you figure out what it objects to or not, even if you have no ability to put the intuition into words. You can treat an intuition like a black box which takes inputs (ideas, like proposed actions or beliefs) and gives outputs (how good or bad it thinks that is, and also how much it cares which might be zero). Non-verbal intuitions can be loosely characterized as mapping all ideas to two numbers: goodness and relevance. Goodness is on a scale that also includes badness (probably as negative numbers). Relevance means how relevant it is to the intuition – how strong an opinion the intuition has about that issue. By checking what your intuition says about many ideas, you can gather a bunch of data points and get a sense of what the intuition wants and cares about.
You can also consider where you got an intuition and whether knowledge from that source is something that you have much respect for. For example, you may have an anti-homosexual intuition from going to church as a child, but you no longer respect that church. That doesn’t mean you should just ignore the intuition or hate part of yourself though. It’s better to face your situation honestly and work on untangling your ideas. Self-improvement is a gradual, ongoing journey, and pretending to be better than you are can sabotage it. If you try to suppress an anti-homosexual intuition, one result is you won’t learn about the issues more thoroughly. By trying to move on by rejecting the anti-homosexual ideas that you consciously think are bad right now, today, you’re actually preventing finding out that some other ideas you have are anti-homosexual. You don’t recognize them as a problem yet, but you also think you already know the right answers and you’re already done changing. You’re pretending to be done learning when you aren’t, which helps protect some mistakes from being changed. It’s also important not to assume the conclusion; sometimes your intuition is correct even though your explicit reasoning disagrees with it. You can’t really know what the right answer is until you rationally resolve the conflict between the ideas. It’s like how you can’t know what idea will win a debate until the debate is over.
You can also consider whether the source of an intuition is applicable. For example, you might have a strong intuition from playing poker that isn’t actually relevant for chess, or vice versa. Even in that case, just ignoring your intuition is not a good solution. There’s an error in your knowledge. Your intuitive knowledge is mistaken about what categories it applies to. E.g. while playing poker you learned something about all games and made that knowledge intuitive for you, but actually it only should apply to games with randomness, so what you learned was mistaken. It’d be best to correct your idea so that the intuition stops coming up while you play chess. If you struggle to change your idea, that generally means there’s something about it which you don’t understand, which could mean it’s correct in a way that you don’t realize and were trying to incorrectly discard. When intuitions resist going away, even though you think they’re wrong, that often is because they’re right about something that you don’t understand.