| > You > "But there simply is no such thing as "the reliability of the claim" that is distinct from "the reliability of the method used to arrive at the claim...". > Also you (in response to my question "And claims that are correct descriptions of reality are not reliable "_for_" such use?")> "Oh, sure, of course they are!" I'm not sure what you are getting at?! If you think I should be objecting to you using the word "correct", because that is a property of the claim itself and not dependent on the method used to arrive at it: Sure, assume that I object to that. But it seemed to me that that wasn't relevant to what you were actually asking about, namely, whether reliability of a claim implies potential usefulness of that claim _for_ some further purpose. > Would you provide a source on this being the common usage? I really don't understand why you are asking this. I already explained why the notion of reliability of the claim itself doesn't make any sense/is a category error. And I suppose that you would agree that people do use phrases like "claim X is reliable", and that it is usually perfectly understandable what they mean by that. Also, under the interpretation that I suggested, you generally do understand what people are trying to tell you when they use such a phrase, while avoiding the obvious logical nonsense that follows if you assume that reliability is a property of the claim itself. Now, it may well be true that many people who use such a phrase have never really thought through what exactly they mean by that, and thus have never made a conscious decision of whether they are using one definition or the other. But that isn't a reason to think that they likely intend to use a definition that leads to contradictions. It seems like you are asking me to provide a source for the claim that people are not generally intending to talk nonsense. It's not that that isn't a thing that can be investigated empirically, but I doubt that you actually disagree on this. > If that is the case, it is not obvious, because for every claim "X" there is also the claim "not X", meaning that exactly half all claims are true. So you'll need to establish what would tip that balance. Well, while that is kinda true, it's also not really particularly relevant in this context? Even if there were a 50/50 chance, that would not be a useful source of information, would it? The other problem with that perspective is that it counts redundant claims more than once. Like, if some measurement of X could give the result 0, 1, or 2, then "X is not 0" is the same claim as "X is 1 or 2", just phrased differently. So, if you want to get a useful result, you have to rather compare non-redundant claims of the same structure and at the same resolution. If you do that, you notice that there in particular are tons of claims about numerical values, and that there is an infinite number of either discrete values of or equal-sized intervals that you could plug into any such claim to make up new claims--but for any particular measurement, only one of those infinitely many possible values or ranges will correspond to reality, and almost all claims that you could make up this way will be wrong. If you don't find that convincing, I guess we might also replace "made-up claims are almost all false" with "made-up claims are no more likely to be true than chance", as you seem to already be agreeing with that? I don't think I drew any conclusions that require the stronger claim, so we might as well go with a lower lower bound instead. > Let A be the true claim "this bridge can bear 1632 tons", and the method that produced it is "making the claim A up on the spot". > Would you still say these phrases are equal?: "the claim A is reliable" = "it is likely the claim A is true" = "making the claim A up on the spot to establish claim A is reliable"? Well, they all evaluate to false, so yes? Though I don't understand why you mention that the claim is true, as that makes no difference!? In the scenario you are describing, there is no reason for anyone to believe that it is true, because all they know is that someone made it up. > Are you asking me? No, I am telling you--though I wondered why you asked the original question, as the possibility to falsify that claim seems kinda ovious!? > In other words, by your definitions, there is no way to establish with certainty that a claim is not falsifiable, if its not being falsifiable can be falsified. So? > Reliable for detecting the reality that they are detecting. Then I don't even understand what the point of that question is. Like, what else would our senses detect, if not what they detect? Could we maybe transfer this to a different scenario where it isn't as easy to get confused by intuitions? Like, assume there exists a robot with a bunch of sensors. And let's assume you can remote-control this robot, and remotely read what its sensor inputs measure. Now, with this robot you could do what I described above: You could check all those sensors for consistency. And maybe you could somehow communicate with other robots and check for consistency with the readings of their sensors. None of that guarantees that the robot (and thus you through it) sees "real reality", for all you know, it was built with all sensor connectors linked up to a computer that generates fake sensor signals from a simulation running on that computer, including a simulation of the stimuli from interacting with other (actually simulated) robots. But you said that that wasn't what we were talking about. So, if we aren't talking about that, then we are talking about actual sensors!? Is your question whether this guarantees that the sensors are free of any defects? (Well, no?) Or whether it is guaranteed that the robot itself can not be defective? (No?) Or ... what else? I really don't understand what you are asking here. |
I have never stated that I thought this. You may have thought my wording implied that, but that has never been what I was getting at. You've incorrectly inferred this on your own, so I never addressed any statement you made regarding it as it was getting away from the point. But clearly since you've struggled to wrap your head around that, I've now spelled it out for you.
> It seems like you are asking me to provide a source for the claim that people are not generally intending to talk nonsense. It's not that that isn't a thing that can be investigated empirically, but I doubt that you actually disagree on this.
No, I've been getting at asking why you think people mean "the method used to arrive at claim X is reliable" rather than something like "claim X can be trusted for use", when "claim X is reliable" is stated.
Your stance is also noticeably less coherent when it is considered that "claim X" can't be replaced by "this wrench" or "this car" and not lack significant ambiguity ("the method used to arrive at this car is reliable"). Where as if you were to replace "claim X" in "claim X can be trusted for use" for "this car", you'd get "this car can be trusted for use".
What you've done is equivocate, using "reliability" differently for claims versus how it is used in all other areas. In particular using it in reference to X on one hand and in reference to the method that derived X on the other, which has nothing like a mechanical analog to help indicate that your usage is one that matches the usage of anyone else outside of your own head.
> Even if there were a 50/50 chance, that would not be a useful source of information, would it?
Then perhaps what you mean is: Made-up claims have a greater probability of being untrue than claims that are not made up. If that is the case are you asserting that "Any method that doesn't make up claims has a greater probability of being true than any methods that do make up claims"?
> Well, they all evaluate to false, so yes?
So, you think it is possible for some claim to be true and it not be likely that said claim is true?
> So?
According to you: The assertion of any claim that is not falsifiable is not warranted.
This has no certain basis, by your definitions. As a result this means that no claim should be ignored just because it is currently considered to not be falsifiable.
Another question comes to mind. Say you claim "water only boils at 150°C+ at 1 atm". You then perform a test on water at "1 atm", and it boils at 100°C. Having falsified the original claim, according to you, you can now say "it is not the case that water only boils at 150°C+ at 1 atm".
After falsifying the claim "water only boils at 150°C+ at 1 atm", is the claim "it is not the case that water only boils at 150°C+ at 1 atm" falsifiable, and if so how might it be falsified?
> Then I don't even understand what the point of that question is. Like, what else would our senses detect, if not what they detect?
That's partly my point, they wouldn't, and the other part is to have you recognize that there is a self-evidently true claim that is not falsifiable. You can't falsify the claim "our senses have at least one way in which they are reliable". To say that you could would be to say you have the potential to reliably, with your senses, demonstrate that the claim "our senses have at least one way in which they are reliable" is false.