Philosophy of Science 2 - The Hypothetico-Deductive Method

Рет қаралды 14,895

Күн бұрын

Пікірлер: 20

@hannsjurgenhodann6268 5 жыл бұрын

Thank you for elaborating this series of videos. I find the videos very interesting and helpful, and I particularly appreciate the clear and understandable explanations they provide.

@DontSee-lw2jv Ай бұрын

26:32 The observation "All non-black cars are non-ravens" confirms the hypothesis "All non-black things are non-ravens". Why is it the case? Because "all cars are things" doesn't mean "all things are cars". Even if we observe that all non-black cars are non-ravens, it is possible that we might observe non-black non-car things that are ravens. So, it is very uncogent to claim that. Would any scientists who use the hypothetico-deduction method claim that? If they don't, then that objection can't be raised to the method.

@Ansatz66 8 жыл бұрын

The hypothetico-deductive method is also deductive in another way, since it doesn't _only_ talk about confirmation. It also says, "if [the observations] contradict the prediction, the hypothesis is disconfirmed." That's a deduction, and it's the more interesting part of the method. The confirmation part of the hypothetico-deductive method is very vague and useless. What exactly does it mean when "the hypothesis is rendered more probable"? By what measure of probability is it more probable? How much more probable? It seems to be just weasel words, and almost all the objections given in the video come from this problem. The hypothetico-deductive method surely works best if we focus purely on the deductive parts and ignore all this awkward confirmation stuff. The confirmation never really amounts to anything anyway, since all it tells us is that something is more probable in some unspecified way.

@davidlogan8905 8 жыл бұрын

Whether the observations confirm or disconfirm our hypotheses is out of our control, so when an observation does indeed disconfirm the hypotheses, we can't say 'oh we're using the deductive aspect here', we say we're using the hypothetico-deductive method to affirm something which has consequently been denied by our observations. Sure, it only tells us that something is more probable in some unspecified way, but surely that's our cue to specify it further and narrow down the possibilities. It's not as if every single aspect of a hypothesis is going to be contradicted, and it seems odd to reject all of those confirmed aspects and not use it to feed the next hypothesis. What would that amount to? Maybe I'm not understanding you properly here, but it sounds like you're advocating a seriously roundabout way to the same thing. You want the method to amount to something (something that is true, presumably), so why go through the tiring process of deductively disconfirming everything else in order to arrive at that truth instead of allowing ourselves to be guided by the little (and probably very useful) truths that are revealed to us through our investigations?

@Ansatz66 8 жыл бұрын

David Logan "It only tells us that something is more probable in some unspecified way, but surely that's our cue to specify it further and narrow down the possibilities." The problem is that it _doesn't_ tell us that something is more probable in some unspecified way. The reason the hypothetico-deductive method is so vague about how things become more probable is because things don't actually become more probable and if we tried to be specific then it would be even more obvious. We can't say how much more probable a hypothesis becomes because there is nothing to say. Just consider the problem of irrelevant conjunctions. That alone makes it obvious that hypothetico-deductive confirmations actually do nothing to make a hypothesis more probable. "It seems odd to reject all of those confirmed aspects and not use it to feed the next hypothesis." One thing that the hypothetico-deductive method gets right is allowing a hypothesis to come from anywhere. It doesn't matter whether a hypothesis comes from a previous confirmation or from a dream; all that matters is how the hypothesis is tested. "Why go through the tiring process of deductively disconfirming everything else in order to arrive at that truth instead of allowing ourselves to be guided by the little (and probably very useful) truths that are revealed to us through our investigations?" Because we have no reason to believe that confirmations actually reveal truth. Finding a million white swans doesn't reveal that all swans are white. If we seriously want the truth then we'd better work for it, no matter how tiring it may be.

@reda29100 Жыл бұрын

20:43 I'm not a mathematician nor a science philosopher, but a thinker nonetheless. As I recall the truth table, what they mean: (T->T is T) if condition is true, the result will *necessarily* be true (T->F is F) if proposed condition is true and result is false, that can't happen. (F->F is T) it's possible the condition is false and the result doesn't happen. In science context, hypothesis is irrelevant to the observation. (F->T is T) it's possible for a condition to be false and still have the result happen. In other words, in science context, we don't know reality before hand, so can't presume the hypothesis (nor the theories we use today as they're post observation, not the decompiled code of the universe). So when we form a hypothesis, all the logic tells us is the falsehood of a theory/hypothesis (because as we've seen, hypothesis could be false yet still have predicted observations matching reality, i.e., F->T is T), and can't really confirm it as in beyond a shadow of a doubt. Think of an abstract condition, then think of the hypothesis (Condition, C=if it rains then, Result, R = I'm going to stay). (C=T, R=T, T->T is T) It rained and I stayed: was the hypothesis violated? No. (C=T, R=F, T->F is F) It rained and I went to work: was the hypothesis violated? Yes. According to the hypothesis, you should've stayed at home. (C=F, R=T, F->T is T) It didn't rain and I stayed: was the hypothesis violated? No. The hypothesis says nothing about whether the result can or can't happen if condition isn't met. Either results (staying or going to work) are technically possible under this hypothesis. (C=F, R=F, F->F is T) It didn't rain and I went to work: was the hypothesis violated? No. Similarly to above case, violation of condition ctells us nothing (i.e., gives us no prediction) of whether the result would happen or not. In other words, if conditions weren't met, we can't make any predictions to test to either confirm or disprove a hypothesis.

@niket527 2 жыл бұрын

When you say that a "hypothesis is confirmed" (for ex, around 29:20) do you mean that it is conditionally confirmed? As in "this observation makes it more likely that my hypothesis is correct?" Because there's always the chance that a later observation could disprove your hypothesis.

@no_special_person 3 жыл бұрын

Kane b, the best youtube channel

@MatthewMcVeagh 7 жыл бұрын

Not sure we should ever have more 'confidence' in a theory except in practical terms, e.g. engineering or medicine. There's no value to greater 'confidence' mentally except a weak psychological need which undermines the open-minded quest for knowledge. I'd also, and relatedly, question the 'legitimacy' (or illegitimacy) of a belief - what does this consist in, and how do we measure it?

@Rahel_Rashid 7 жыл бұрын

well done

@omerefeoglu1566 3 ай бұрын

Is there any chance to share slides with us?

@alexmeyer7986 8 жыл бұрын

"science raises the probability of cancer" (Objections, 1.) - oh, I hope not. I was actually planning on becoming a physicist :D

@MatthewMcVeagh 7 жыл бұрын

Perhaps we could replace the word 'smoking' with 'science' in other contexts - e.g. "smoking makes you look cool" or "I'm sorry to inform you, but I've taken up smoking".

@reda29100 Жыл бұрын

First, I know it's a joke. Second, technically, 3:55, for other topics we'd had no idea were related, we could say yes, science does raise probability (1) of asbestos causing lung cancer. (1) I don't know statistical terms well but I don't mean ontological probability that we made the chance more likely (not sure of the term again), but that use/application of science is not science, it's technology. Science is just the knowledge, model, laws and formulae. Putting those information to use is technology, not science in strict terms. But revealing facts that cause us to go from (we have 0 reason to think asbestos causes lung cancer) to (we think it has %20 chance or whatever or causing lung cancer to those exposed to it for X number of hours for a month or so). That epistemological probability meaning increasing is what I call (science increased probability of smoking), cuz technically it did make us think smoking is more likely to cause lung cancer than we thought before (that it has no role to play).