Many people's favorite argument lately is to assert that something has "no evidence" for it. In general, there is lots of evidence, which they don't count for various reasons they usually don't explain. The kind of person to say this doesn't usually make a list of all the evidence and then go point by point and say why each piece of evidence counts as "no" evidence.
What do they dismiss? Generally anything that's merely similar but not 100% direct proof. They want the exact thing in question to have been directly tested. For example, there could be dozens of studies showing that masks work with dozens of different viruses. But then when a new virus, COVID, comes along, some people will say there is "no evidence that masks work for COVID".
They should accept that the evidence about masks and viruses applies to COVID because COVID is a virus. Instead, they focus on the lack of studies specifically testing masks for COVID.
And evidence is broadly overrated anyway. Lots of issues should be considered with conceptual thinking, logical reasoning, explanations, and critical arguments. Evidence is most useful when you have multiple ideas which contradict each other that you don't know how to choose between. If you fail to resolve the issue using many other more convenient tools, then you look for differentiating evidence. Before that point, you don't know which evidence to look for.
The concept of looking for evidence for (or against) one idea is wrong. The only way that can work is if you find evidence directly contradicting the idea, in which case you can rule the idea out. Other than that, searching for evidence about one idea doesn't make sense. Which evidence should you look for? Why that? There are no good answers.
Instead, it makes more sense to look for evidence when considering two or more ideas. Which evidence do you look for? Look at issues where the ideas make different predictions about the same thing. Then observe and see which idea (or neither) was right. They can't both be right when they make different claims about the same thing.
Having two or more ideas tells you where to focus: you can specifically seek evidence that'll differentiate between those ideas. You figure out what they disagree about, which is related to the physical world, and then go look at or test the world to get a result. Then, since the ideas disagreed, no matter what result you get, at least one of them is wrong, so you can rule something out. There's still no such thing as positive evidence supporting ideas. All you can have is evidence that contradicts ideas or does not contradict them. Non-contradiction is also called consistency, but that isn't support or justification.
The idea of looking for evidence to differentiate between two or more ideas, rather than to support or evaluate one idea by itself, comes from Karl Popper. So does the idea that evidence can refute ideas but can't support them.
Say, as a simple example, that I'm thinking of an integer from 1 to 10. John thinks it's 7. Alice thinks it's 4. They can't both be right. If I reveal my number, at least one of their predictions will be refuted.
Say they both claim to have telepathy and that my number is 4. That is negative evidence about John's telepathy. He was wrong. Is it positive evidence about Alice's telepathy? No. This particular piece of evidence doesn't refute Alice being telepathic, but it doesn't positively support or justify it either.
In general, is it possible to test the exact right scenario? No. You can do a test of masks with COVID, but you'll do it in a different room than I'm in. How do you know it works in my room, or outdoors, rather than in the room you ran the tests in? There is no evidence – no published, peer-reviewed studies – that it works at my location or in the identical type of surroundings to mine.
You must always use conceptual/creative/logical thinking, not just evidence, to decide how to apply results. If it works in Idaho, will it also work in Texas? Yes, because the physics of air, viruses and masks is the same in both places. Should I also wear masks on the moon? Not necessarily: the moon doesn't have atmosphere, so there's a relevant difference there and reconsideration and possibly additional evidence is required.
A scientist may test 10 masks. But my mask isn't one of those ten. It may be the same type as one of those ten, or it may be a similar type, but it's certainly not one of the masks that was tested. So there is "no evidence" that my mask will work because my mask has never been tested. This is absurd, but arguments that are no better are routinely made in public, by respected intellectuals, and accepted by audiences. Usually they are made in favor of conclusions the speaker, and a lot of their audience, are already biased in favor of.
Why are they biased in favor of their conclusion? Some reason other than evidence. They analyze it in some other way, reach a conclusion, then claim the reason they reached it is based on evidence, even though it isn't. They do this because "evidence" is widely considered powerful and considered the appropriate way to evaluate ideas. Arguments and reasoning are less fashionable.
In our culture, saying "all the evidence supports my side" is a great appeal to authority; if people believe your assertion is true, they'll be really impressed. But if you say "all the arguments support my side" people aren't as impressed. It doesn't work as well. People seem to think that confusion about arguments is common – arguments are fallible – but evidence is more safe, reliable, trustworthy. Of course, sometimes they're wrong, at which point they may say "the evidence changed, and I changed my mind with it" or similar. They do know the current state of the evidence can change, just as the current state of the arguments can.
I think appeals to authority are bad. But they do seem pretty popular. I think our society would benefit a lot from better rationality.
Perhaps the fundamental reason people get away with saying there is "no evidence" for conclusions they dislike is because there is no such thing as supporting evidence. They may dislike and disagree with Karl Popper, but they're also sort of implicitly weaponizing one of his major claims (about the logical difficulties with supporting evidence) against everything they dislike. But they're also hypocritical. They don't use the same standards for ideas they do like. For their own ideas, they claim there's lots of positive, supporting evidence.
The correct way to judge ideas is: can we come up with any alternatives which are not refuted by evidence or critical reasoning? If you use criticism, including empirical (evidence-based) and non-empirical criticism, to refute all the alternatives anyone has come up with, then you can reach a conclusion because you have one single idea left that survived criticism. At least that's what I think. Popper wrote a lot of things which kind of suggest a viewpoint like that, but he didn't get all the way there. He said you could use a bunch of criticism and then evaluate which idea best survives criticism and use that even if there are several contradictory survivors. Popper didn't focus on decisive arguments like Critical Fallibilism. I think Popper's proposal makes more sense than the standard view. It's important to use negative arguments and to consider alternatives. If you want an idea to win, you need to figure out and criticize alternatives, not ignore alternatives and add a bunch of support to it. Support doesn't work, both logically and practically.
Note as a detail: People commonly bring up the concern that alternatives haven't been researched enough. But "this is under-researched and more research is needed" is an alternative that's very easy to think of; it's generic and you can repeat it for many subjects without having to creatively innovate each time. So that is one of the things you should consider, and if all alternatives are refuted it means you have a refutation of that.
A better thing to say instead of "there is no evidence for X" is instead "the evidence for X isn't a perfect fit". In other words, none of our evidence directly confirms X as a really specific, narrow claim. Instead, our evidence shows patterns that fit X pretty well but don't fit every detail. That doesn't mean the patterns contradict any details, just that some details are omitted.
For example, the evidence that masks help prevent the spread of the flu isn't a perfect fit for whether this mask, in this building, will help prevent me from getting COVID.
If you say it this way, it leads to discussion about what patterns we're aware of and have evidence for, and how the patterns should be extrapolated or generalized. What cases are similar enough that the evidence for the pattern is evidence for that case, and when does it not count or not apply? That would be an honest conversation, and potentially productive, even though I think that kind of positive-evidence-for epistemology is incorrect.
But when people say "no evidence for X" they're suggesting there's nothing to talk about. They're suggesting there are no patterns that may or may not fit well enough to count and nothing to be discussed. They're suggesting there is no evidence/pattern that is even close. They're being super dismissive instead of opening the door to conversation. That's usually dishonest. If it's some dumb issue like "no evidence that aliens visited us and planted fake dinosaur fossils" then whatever, the phrase isn't such a big deal. But for reasonable topics where some reasonable people are on the other side, the people saying there is "no evidence" for the other side are just trying to prevent debate by being unreasonably dismissive. They're treating the other side like they're being as reasonable and evidence-based as people claiming aliens planted fossils, even though they're just saying masks work for COVID. That shows poor integrity.
Also, when people say "no evidence for X" it often sounds to the audience like X was studied and failed. But the real situation is often that there's no evidence just because it wasn't studied yet, so if it were studied you don't know what the outcome would be.
Also, as above, often some relevant things have been studied, and it takes thoughtful discussion to figure out if they are close enough, if their patterns should be extrapolated to this case, or not. Or put another way, even if the patterns are a pretty poor fit, wouldn't that be a small amount of evidence, not literally none? Stuff has to be pretty dumb for "no" evidence to be a reasonable, honest claim instead of "low" evidence. So "no evidence" is generally a red flag about the speaker, whereas "low evidence" is a hint they might be a more precise, reasonable, honest person who is open to conversation instead of really dismissive of alternative ideas.
Tons of stuff hasn't ever been studied well, even things that many people claim, assume or expect was studied well. So "no evidence" doesn't really mean as much as it sounds like. Like there's often "no evidence" that a vitamin helps with a particular disease because that hasn't ever been studied. Medical studies on humans are expensive! But there certainly is evidence that that vitamin has various benefits, and there's some evidence and knowledge about its function that lets us reason about what diseases it could help with, so there is absolutely relevant evidence that could be extrapolated to at least partially apply even if the exact issue hasn't gotten an expensive study.
Messages