Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!
You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!
What if I say something that is logically falsifiable, but not empirical falsifiable and that produces something I find useful, couldn't that count?
Don't try to play word games! Scientific Theories are things which are subjected to falsification, not conversations. What on earth would a conversation "within the realm of empirical science" even look like?! It's just nonsense. You're talking cleverly worded gibberish that at a first glance sounds meaningful.
I think you're missing the point.
Define vacuous prediction.
How do you know this?
See, I don't understand what you mean by real. My theory makes a prediction. You say this isn't "real" because there is some other theory that makes the same prediction?
Massive lolz if you think I was using that seriously
someone obviously doesn't follow all those God threads haha
See I find this hugely ironic, because you've actually been more influenced by philosophy than me.
You would compare the Baye’s factors.
You should stop when you say “if we assume that.” Show me that the assumption is correct before you continue. Science is about testing our assumptions, not about letting them go unexamined.
At first glance, I found that it was verified almost two decades after relativity was accepted, but my source could have been in error. Do you have a source for when the speed of light being constant was tested? However, this is irrelevant since the “counterexample” of Mercury at the time shows that your claim, that one counterexample is sufficient to overturn or invalidate a theory, is false.
The point of science is to test your ideas, and that includes your models. If you include what you know into the model, which you would have to in order to make the model consistent with the available data, you can’t then test the model to see if it is accurate against that same data. What you have just shown is that the geocentric theory did not make any successful predictions, making it not scientific. I think you are proving Alt4’s point that you don’t know what science is.
Science is the process of testing your hypotheses, not about seeing if they are consistent with the evidence. If the latter were true, then science would be flooded with ad hoc explanations. However, that is not the case. Many statements can be made about an unfalsifiable theory. Because it makes no predictions, it can’t be tested, which is the method that we use to evaluate our ideas. It follows that ideas that go through this process are more probable and therefore more preferable than those that have not. Just because two theories are consistent with the facts does not mean that they are on equal footing. This merely supports the parsimonious principle that we should avoid ad hoc explanations. Making your theory fit the data after viewing the data makes it ad hoc and therefore needs to be tested in order for it to have any credibility.
Yes, a form of utilitarianism. I’m not sure how much of a debate one could have other than semantics. You could also challenge the philosophical underpinnings of geology in the same fashion. This is exactly what I said earlier, that there is no more philosophical debate to be had that doesn’t apply to any other field of science. I don’t find this debate on semantics to be very interesting, which is as far as I can see is the philosophical debate between moral theories.
Science completely fails your criteria here.
That's very perplexing. You don't think that for instance, physicists discuss their findings with each other? What about science class? What about any scientific paper? Discussions within empirical science occur all the time, regarding new findings and interpretations and the like. Scientific findings must be communicable obviously, this is what I mean by conversing within the realm of empirical science. In this conversation, almost none of our claims have been scientific in nature, including the one I'm currently making. Is this all useless? Your scientism makes even discussing these issues impossible.
If all you mean by that is descriptive utilitarianism, then I'd agree, debates on descriptive morality can only be semantical. But if you're advocating this prescriptively, then I'd maintain that the debate is not semantical and that utilitarianism makes no objective prescriptions.
Science
Ok, I'll stop trying to be coy. I think philosophical assumptions go into these.
Standard Bayesian practice. How else do you determine the initial prior?
I read the book "Relativity" by Einstein, and it talked about how there were things showing the speed of light to be constant (something related to Maxwell's laws maybe?) - which led him to his ideas. Very philosophical book there, by the way. Einstein makes the philosophical assumption that the principle of relativity (laws of physics are the same in any frame of reference) holds, and then determines what he can from that.
If a counterexample exists, then you know your model is wrong. If I have a model that says when I drop things, they fall, then a helium balloon shows the model is wrong and that we have to change it. Counterexamples are failed predictions.
What are you talking about? First of all, the geocentric theory is just as scientific as any other theory. It was eventually replaced by a theory that matched up better with the data, but that doesn't make it unscientific.
I'm not even really sure what you're referring to here - how did I show that the geocentric theory makes no successful predictions?Wikipedia said:
Except that everything has already happened by the time that you do your comparison.
PHILOSOPHY ALERT!
Yes. But two that are tested and pass all those tests are going to be equally good then, right?
... and what exactly does the phrase "testing your hypotheses" mean if it doesn't mean "ensuring they are consistent with the evidence"?
PHILOSOPHY ALERT!
Pretty much your entire way of thinking.
I do not. I don't find the concept of "objective prescriptions to be coherent. The is-ought gap would need to be solved first. Until then, I wouldn't adopt that concept.underdog said:
philosophy = search for knowledge. Kinda hard to describe in more specific terms, but there are very diverse branches like ontology, epistemology, etc.
Evaluate the reasoning or evidence behind it?
Strictly speaking yes. I think of mathematics as a subset of logic, and logic is at the least an extremely important part of philosophy.
No. There are some people who support mathematical empiricism, but I disagree with them. I think of mathematics as statements of implications - so the axioms of ZFC imply Banach's fixed point theorem. I don't worry about justifications for the axioms.
I would explain what 2 means (the 1st successor of 1), what +2 means (apply the successor function, then apply the successor function again), what equals means (is the same as) and what 4 means (the 3rd successor of 1) and show that based on the definitions of the objects involved, the statement is true. It's just shorthand for a bunch of logic.
Wikipedia said:
Yes I have. Correct. This is the way it works. It's not just my opinion of how it works, it's not just my point of view. This is how science works.
That does not follow. You can make predictions about the future easily using only observations. It's basic pattern recognition. Let me give you a trivial example:
Correct again. Science does assume "the future will be like the past". Science makes many assumptions we cannot prove. It deals in the real world, not academic fantasy.
So by "conversation within the realm of empirical science", you just mean a discussion involving evidence and how well they fit theories? I'd love to have that conversation, here. That's what I've been trying to accomplish all along.
No, being inconsistent with the current data negatively effects the prior probability. I'm saying that when first proposed, models have a low prior probability and in order to increase our confidence in them (the epistemic probability of the model), they must make successful predictions. This is why unfalsifiable models are not very good, because they have a low prior probability and are unable, by definition, to increase our confidence in them. Fitting past data helps, but it is not a requirement. It depends on our confidence of the data that the model does not predict. If we have a high degree of confidence in the data, then the model probably won't overcome it. If we have a low degree of confidence in the data, then successful predictions by the model might make us reconsider the previous data (i.e. perhaps it is an illusion, maybe it was a hoax/human error, taken with faulty instrumentation, etc.). Back to the original question, the opinion that we would have of it is that it is a remarkably improbable idea and until it passes an extreme amount of tests, ones that show (large parts of) our current knowledge to be incorrect, the idea has no chance of being taken as credible.ballin said:
Science is a subset of philosophy. In fact, for a LONG time "science" was actually known as "natural philosophy".
Ok, lower prior is one way to look at it - but that prior is really just another series of Bayesian estimations.
The reason why I state things like I do is to avoid several biases. For example, hindsight bias could become a problem if people only used what is already known to test their ideas. They could always say, my theory predicts X, Y, and Z ( where X, Y, and Z are things that have already happened) when in reality, their theory does not predict X, Y, and Z. This is common in religious discussions. People say that theism predicts that there will be a universe with life that can do calculus and an omnipotent being that wants to console our suffering by causing a demigod to suffer. Right, so a universe with no suffering would be evidence against the existence of God? No, its just that they are fitting their theory to the data, not making predictions based on their theory. You can't say that P and not P is evidence for your idea. However, this doesn't prevent someone from observing the outcome, either P or not P, and then claiming that the observed outcome was evidence for their idea. When you make predictions concerning already known observations, you become vulnerable to this bias. If you claim your prediction before the observation and it fails, you can't then retreat and say that the observation fits your prediction. So, how do you distinguish between someone who is biased in this fashion and someone who is not? The biased person can only successfully predict things that have already happened, and the non-biased person can successfully predict future events. This is why the prediction of future events is important, it makes the hypothesis that the model fits the data preferable over the hypothesis that the model was fitted to the data.ballin said:
Yes, and words change over time. The meaning of words is justified by their current usage, not the usage of two centuries ago. Nobody from that time is going to benefit by keeping the same naming conventions, yet many people from today will become confused if you try to revert back to their naming convention.ballin said:
A theory should test against all available evidence. As more evidence comes to light, of course that gives more information about the theory.
It's all in the way you look at it - if you believe in hard determinism then the pattern simply is the pattern (not sure if that makes sense).
Priors matter too though, a LOT. If the pattern is "my iPhone falls when I drop it", I don't think we actually have to test that to be pretty confident in it.
I'd think many would agree with me though as to the status of science as a subset of philosophy.
I'm not so sure about that. If you look at how the term scientist was coined, you will see that they were meant to be mutually exclusive. It was coined because philosophy was used to refer to "armchair philosophy." I don't think that this sentiment has changed, especially among scientists. You wouldn't get so many scientists disparaging philosophy if they considered themselves philosophers.ballin said:
I think you are confusing the prior and posterior probability parameters. For the example of gravity, the prior would be the probability of gravity is true before we saw the effects of gravity. This would basically be asking, if I were to drop you into a random universe, what is the probability that things with mass attract each other? We don't know if this is common or rare, it may be more common that things with mass repel each other, but that we can never observe such a universe because living things can't exist in it. The question you are asking, what is the probability that things with mass attract each other after we continuously observe things with mass attract each other relies on the posterior probability, not the priors. I think that gravity has a low prior probability and an extremely high posterior probability.
So is every starting condition for the universe going to produce background radiation? Is every starting condition going to produce gravitational forces? Is every starting condition going to produce X? If not, then you can find the implications of those starting conditions and test those by predicting the existence of a phenomena that was not already known at the time. They may not be testable at the moment, but I think that it is an overstatement to say that they will never be testable.
Gay bashing slander? What are you talking about?
Is there something wrong with sucking ****, Dre? And you wonder why everyone thinks you're a homophobe. Maybe it's because you say homophobic things?
I don't see scientists disparaging philosophy. No matter what anyone says, science is based in philosophy - (philosophy of science).
No, now you are changing definitions on me. You said that all past evidence gets included in your prior. I said that's fine, because that's one way to look at it, and I'm reusing your definition here. I usually use the term prior in the sense that you do here - as the probability without any evidence, but you used the term prior to mean probability given all the past evidence.
The question is whether I need to do additional tests for my theory that my iPhone will fall when I drop it - which I have formulated based on the fact that it has fallen when I have dropped it in the past. I don't think I need to do additional tests, but you apparently would say that this is an unacceptable ad hoc explanation.
Unless we can run an experiment to create a universe it will be pretty tough.
How can you say that is true? Their is no truth on your view, even making a statement about somethings usefulness is making a truth claim. You cannot escape making truth claims, which is why you've been doing it this entire
Not really, you would simply have to assume that the past will resemble the future and that the pattern would hold for the future and you've already conceded that science is not capable of proving that and simply must assume it to make predictions. And that is what I mean when I say we have no empirical access to the future, we cannot measure it and it takes non-empirical assumptions to make predictions about it.
I think you get the point. Every single statement that you make is a truth claim. Now if science doesn't deal in any truths, then you either have to stop saying basically anything, or admit that you can have other sources of knowledge other than science.
But that's exactly the point! Science can't access certain areas that philosophy can. And if you won't accept the legitimacy of philosophy, you have to give up all of those areas of knowledge, and science hinges upon many of them. All that a scientismist could conclude is that it must be pure, dumb luck that science succeeds in predicting anything!
The topic of this discussion is inherently beyond the reach of pure science and that's why you've been in the awkward position of having to make a ton of metaphysical claims while denouncing philosophy as useless, let's at least be honest about that.
Stephen Hawking and Leonard Mlodinow remarked in their most recent book that “philosophy is dead” with respect to the physical sciences. Neil Degrasse Tyson has echoed this remark. Richard Feynman is notorious for his denigration of philosophy. This is going to get stale quick since the discussion has devolved into semantics…basically, they would disagree with your definition of philosophy…and they would narrow it to what philosophers do.ballin said:
I misspoke, I mean to say that you have some idea with prior probability X, and in the past it failed its tests (inconsistent with known data), then this negatively affects the probability (which was previously equal to the prior, but now is the posterior probability), so now the probability is less than X. When you start out, the prior probability equals the probability. As evidence comes in, the probability will change based on whether the evidence is favorable or not. The evidence will be positive or negative to this “prior probability.”ballin said:
Since when is gravity ad hoc? It has been extensively tested and has always succeeded. It is a prime example of a model that has been tested so much that our confidence borderlines certainty, which makes it a really good model. All I am saying is that “predicting” the past data is not sufficient. For example, an omnipotent god who makes the world exactly how we just observed it. If we only argued from past data, this hypothesis would technically be confirmed. If the past data turns out somewhat unexpected (it was mistake, hoaxed, etc.), then a different omnipotent god would be the hypothesis. There would be a different god with different vision of how he wants the world to unravel for every possible world with new evidence confirming some god’s and disconfirming others. By predicting future events (whether the prediction was made and (dis)confirmed in the past is irrelevant, as long as the prediction points to the future), you avoid these ad hoc explanations. I think we are on the same page here, so I won’t push this any further.ballin said:
I don’t recognize this principle. We have a theory of gravity that things with mass attract each other. This theory is very well tested so we have a great deal of confidence in it. Your backpack has mass, so it will be attracted to the mass of the Earth. Now, does this generalize to objects that don’t have mass? No. Now, if you are testing a theory that has conditions that generally applies to things and it is a successful theory, then of course it is going to apply to things generally. If it didn’t it wouldn’t be the same model. This has nothing to do with a principle that results generalize, it has to do with the specific hypothesis that you are testing.ballin said:
Maybe they don't understand what philosophers do. Very few philosophers make claims that have to do with science, so I'm not sure why they would have any authority about philosophy in general anyway.
It's all in how you look at it. But it's still true that past evidence matters just as much as future evidence. I mean, future evidence literally becomes past evidence as soon as you know it.
I was using the example under the assumption that we don't already have a theory of gravity. Looking at the fact that things have always fallen in the past, I still think that increases the probability of our theory of gravity being correct even without future tests.
Why do we think that all things with mass attract each other, instead of just things that we have actually observed attracting each other? We have observed that, say, a tennis ball tends to fall when you drop it, and we think that ALL OTHER THINGS WITH MASS tend to fall also. There's really no evidence for this over the theory that ONLY tennis balls fall.
So merely finding a past pattern means that the idea is more likely to be true? I don’t follow your reasoning.ballin said:
You have to look at the Baye’s factors. If the hypothesis is that only tennis balls have a gravitational force, then there is a mountain of disconfirming evidence.
We can test whether helium has a gravitational force. If it didn’t, then there would not be helium nebulae. There are, therefore helium has a gravitational field.ballin said:
Science doesn’t deal with absolute certainty. If that is your goal, then expect to be disappointed. There are only things that we know with varying degrees of certainty, with some that approach but never reach absolute certainty. Your comment about generalizing doesn’t support what you originally said. All your comment says now is that theories that deal with general concepts will apply generally. This is a trivial statement. This side conversion started with you saying that science has a principle of generalizing results, with no conditional attached.ballin said:
Yes. That's exactly what I'm saying. If there is a past pattern, it is more likely to be predictive of the future. This is because the future resembles the past.
Perhaps I'm not explaining this clearly. How did we come up with gravity initially? We saw that some objects had a tendency to fall. Ok, now pretend that we only know that tennis balls fall - it's not THAT dissimilar from what really happened. Do you think that other objects are more likely to fall, given that we have done tests that show balls are likely to fall?
Random point: Science does deal with certainty on FALSE things, as I mentioned before.
How so? You can always question the validity of previous evidence. We tend to think that the evidence attained is very good, so it is very, very likely that it will stand, but if you deal with certainty on this front, then you will not be open to evidence of hoaxes, mistakes, faulty equipment, etc. Remember, absolute certainty means that there is nothing, absolutely nothing that will persuade you, including if someone goes “here’s a flaw in your reasoning.” In science, you can approach, but can never reach absolute certainty. To do so would be to claim oneself as incapable of error.
It doesn’t make any predictions. Saying in the future, blah blah blah does not allow the ability of being false because it is consistent with every possible outcome. If you were to give a certain date, then you would be able to test it. Suppose someone proposed that the sun will continue to rise for the next 100,000,000 years, but on the 100,000,001rst year (+ or – some margin of error), the sun will cease to exist? That would be a very improbable claim until we consider the evidence. Well, we do have evidence that all stars eventually die, and that the properties of the sun is not unique, therefore sometime in the future, it will die. We would have to look at the evidence of the lifespan of stars that are similar to our own and it turns out that in about 100,000,000 years (for example), the lifetime of the sun will expire. It also helps that we know a lot about the processes that make up stars and how that process leads to its own demise. We can test our theory of the stages a star goes through by observing other stars.
The way I have heard the term before usually includes private information. I would limit the scope of this to include only publicly verifiable information so as to reduce the ability of the bias of individuals to creep into the process. This would make it more “objective,” in the sense that the grass is green and the sky is blue. We will still be vulnerable to human bias, but there is only so much we can do to limit that. However, this only means that we shouldn’t rely so much on humans as data information gatherers and build instruments that aren’t affected by moods, which is exactly what science is doing.
We can say with certainty that Newtonian mechanics are not 100% correct, because we have made observations that were inconsistent with Newtonian mechanics. That's what I mean. When you falsify something, you claim its opposite to be true. I agree that theories are never shown to be true, but sometimes they are shown to be false.
Does gravity make any predictions? This theory makes the SAME predictions as gravity.
No, it's consistent with every possible outcome where gravity also holds - this is very different from every possible outcome. Unless you are claiming that gravity is the absolute truth.
The evidence for this theory is exactly the same as the evidence for gravity. This idea is not unfalsifiable or untestable btw - it is tested exactly the same way we test gravity. There is no evidence for this theory over gravity, and there is no evidence for gravity over this theory. Both make the same predictions for all intents and purposes.
Yes, just like gravity. What makes gravity different from this theory?
I'm not sure what this all has to do with what I said. In fact, according to your reasoning above, you have no reason to believe gravity over my theory, because you can't do any tests.
For Newtonian mechanics, by the inverse square law. For relativity, it would be a different equation. Based on the present location of objects, you use the equations to project the future paths of physical objects and if the objects are where you predict they will be, the prediction was successful. You seem to be saying that your hypothesis would predict that everything would end up resting on the surface of the Earth, which is not the case. Every object that has reached escape velocity would be evidence against your hypothesis, but not evidence against Newtonian mechanics or relativity. Also, because Newtonian mechanics and relativity makes more specific predictions, they would have a greater Bayes factor and therefore more evidence than your hypothesis.ballin said:
Have you ever observed something that turned out to be different than you first imagined? Every observation has a probability attached to it. It’s just that the confidence we have for our observations are reasonably high and every time they are independently verified, the probability that the observations were wrong decreases at an exponential rate. This means that the probability that the observations are wrong and that Newtonian mechanics is exceedingly low, llow enough not to be considered seriously, yet not absolutely zero.ballin said:
Predictions are necessary for a test but not sufficient. In order to test an idea, it must be able to fail the test. If the idea is consistent with every outcome of the designed experiment, then the idea has not been tested. For example, when you propose a causal link X->Y, the only way to show this link to be wrong is to have X->-Y. Therefore, if you experiment with –X, you can’t test the proposed idea because –X->Y and –X->-Y are both consistent with the proposed idea. This is why it needs to be specific what you are proposing in order to see what outcomes are consistent with the idea or not. If they actually propose the same causal link (make the same predictions and have the same evidential support), then you are simply proposing the same thing as Newtonian mechanics or relativity and are simply saying that they are different and your implication that they would predict the opposite to happen in 1,000,001 years is simply false.ballin said:
Here is where the mistake from the previous point shows up. They may both be consistent with the evidence, but one may have passed more tests or more stringent tests, and therefore has more evidence than the other. You have not unpacked the implications of your hypothesis as to describe what tests it has passed. As of now, I would be comfortable to say that the evidence of gravity is more tested than your theory, which translates into having more evidence than your theory.ballin said:
Yes, no. When I say successful prediction, I mean that a prediction was made from the model, it was tested, and it successfully passed the test. By test, I mean that it was possible for the model to fail the test. The harder to pass the test, the more evidence passing the test confers to the hypothesis. It is not clear by the formulation of your hypothesis, that your hypothesis is both distinguishable from relativity and has made any successful predictions that rival the success of relativity.ballin said:
By truth claim I mean maintaining that something is the case. For instance, right now you just maintained that it is the case that science maintains healthy skepticism. That is what I mean by a truth claim. I would distinguish between truth claims and absolute truth claims, the difference being the degree of certainty about the truth claim. I never have contended that a truth claim must be absolute.
Knowledge generally does have a very high connotation of certainty, the most usual definition is justified, true belief. So I agree that there is very little we can truly know. The more reasonable question I'm proposing why believe that there were that many birds? Belief has a lower definition of certainty and all we need here is a good reason to believe. You seem to have automatically assumed that the answer to this is scientific, you seem to say that once we mitigate the issue of knowledge we will immediately be brought to science. The issue for you is that the answer to this question must by necessity be philosophical. How could we scientifically come up with a reason to believe that there were that many birds? You may immediately be thinking, well, because we observed that many birds! But this presumes that verification is a warrant for belief, and as soon as you make that claim, you are making a philosophical, epistemological claim about belief. For there is no way to scientifically prove that verification is the way to come to true beliefs. Furthermore I've argued that verificationism is self-refuting and too restrictive.
But how can you adequately make statements about the "real world" without answering metaphysical and epistemological questions. For instance let me put this to you, right now you're actually a brain in a vat being stimulated to experience everything around you. Or, everything that you observe around you is a veridical observation representing a real, objective reality. Now those two hypothesis predict the same exact events neither of them can be verified to be more real or true than the other. Science by itself here can make absolutely no claim as to the nature of the "real world". Science alone can't make any metaphysical claims as to the nature of the "real world".
How could his debate even possibly be scientific? We've never been discussing empirical data and theories here, this conversation is beyond the scope of pure science.
If you make a statement about the nature of the universe as it really is then this is clearly a metaphysical, and thus philosophical claim. Only a statement about an empirical observation or theory is scientific. Unless you want to argue that these empirical observations and theories are veridical representations of reality, and that again would be philosophical, rather than scientific discussion.
I was trying to simplify the situation a bit by sticking to the simple case of objects falling, but whatever, the point stands in any case. Escape velocity and so forth are all irrelevant to the point of the argument - do you realize this? Modify what I said to be "Newtonian mechanics EXCEPT in *insert untestable case here*". Or "general relativity EXCEPT in solar system X of the Andromeda galaxy, where Newtonian mechanics actually hold". Or whatever. Nitpicking the example is irrelevant.
Well, it's still a statement of truth to say that the probability of Newtonian mechanics being correct is low. I'd also maintain that one of the ideas of Newtonian mechanics is that our observations are correct, but whatever.
But my theory was NOT like this at all, as I pointed out before. Unless you think gravity is consistent with every outcome of the designed experiment too. My theory can fail if a bowling ball fails to fall when you drop it.
I think I addressed this in my previous post ...
I don't follow the last sentence. I am proposing something very similar to relativity or Newtonian mechanics, yes.
You say this, but it's not clear at all how this is the case.
My hypothesis is not distinguishable from relativity in any cases that we can currently test. In all cases that we can test, it has made the same successful predictions (note though that it was certainly possible for it to fail, because it was possible for relativity to fail).
